Lesson 4 of 14 · about 60 minutes · Updated November 15, 2026
Policy Management and the Compliance Program
You'll be able to
- Distinguish policies, standards, procedures and guidelines, and assign each the right approver, owner and review cycle
- Run a policy lifecycle that produces evidence: drafting, approval, publication, attestation, yearly review and retention
- Design an exception and risk acceptance process with approval levels, compensating controls, a register and expiry dates
- Map one tested control to many rules using NIST informative references, NIST IR 8477 and a unified control library, without treating a mapping as proof
- Build a compliance calendar from real deadlines such as SEC Item 1C, the NYDFS April 15 filing, FTC Safeguards reports, PCI DSS assessments and CMMC affirmations
- Apply the Sentencing Guidelines, the Justice Department's compliance evaluation guidance and the lessons of whistleblower and False Claims Act cyber cases to a security program
Start here: the short version
- Written policy is a legal requirement, not a formality. New York's 23 NYCRR 500.3 requires written policies on 15 named topics approved at least once a year by a senior officer or the board; the FTC Safeguards Rule, the federal banking guidelines, HIPAA and PCI DSS all require written programs, and the federal sentencing rules and the Justice Department judge a company's compliance program by whether its written rules are real.1234
- Keep four layers apart. A policy says what the organization requires and why; a standard makes it measurable; a procedure says who does what, step by step; a guideline is advice. Mixing them is why policy binders run to 400 pages that nobody reads.
- Exceptions are part of the program. Every large organization has systems that cannot meet a rule this quarter. The question is whether each gap is written down, rated, covered by other controls, approved by someone with the authority to accept that risk, and given an end date. Idaho state agencies, for example, may get an exemption from a statewide IT policy for no more than 24 months.5
- Map once, test once, report many times. NIST's informative references, its 2024 mapping method (NIST IR 8477) and free crosswalks such as the Secure Controls Framework let one tested control answer many rules. A mapping shows a relationship, not that two rules are the same.678
- Certifications are personal now. New York requires the chief executive and the CISO to sign an annual compliance filing by April 15; defense contractors must have a senior official affirm compliance every year in the Pentagon's SPRS system. False statements in either have cost companies millions.91011
- Insiders are watching the paperwork. Since the Justice Department launched its Civil Cyber-Fraud Initiative on Oct. 6, 2021, False Claims Act cyber cases have been brought by a former chief information officer (Penn State, $1.25 million), members of a university cybersecurity team (Georgia Tech, $875,000) and other employees, who received shares of the recoveries.121314
Rules and standards described: 23 NYCRR Part 500 as amended effective Nov. 1, 2023 (sections 500.3, 500.5, 500.9 and 500.17 read Nov. 15, 2026); FTC Safeguards Rule, 16 CFR 314.4, as last amended Nov. 13, 2023; Interagency Guidelines Establishing Information Security Standards, 12 CFR Part 30, Appendix B; HIPAA Security Rule, 45 CFR Part 164, Subpart C (the Jan. 6, 2025 proposed update was still a proposal on Nov. 15, 2026); SEC Regulation S-K Item 106 (Release 33-11216, 2023); PCI DSS v4.0.1 (June 2024); 32 CFR Part 170 (CMMC, effective Dec. 16, 2024), with CMMC Phase II suspended by the Department of War on July 13, 2026; California Privacy Protection Agency regulations effective Jan. 1, 2026; U.S. Sentencing Guidelines Manual, Chapter 8 (2026 edition; section 8B2.1 unchanged in 2025 and 2026); Justice Department Evaluation of Corporate Compliance Programs (updated Sept. 2024, still the current version on Nov. 15, 2026) and the department-wide Corporate Enforcement Policy of March 10, 2026; NIST CSF 2.0 (Feb. 26, 2024), NIST IR 8477 (Feb. 26, 2024) and NIST SP 1347 (2026); Idaho Technology Authority P1015 and G120, read Nov. 15, 2026. Rules, enforcement policies and standards change; check the regulator's current text before relying on any point.
Before you change anything. Plan changes to production systems, test them first, keep a way back, and follow your change management process. For clinical, plant-floor or other safety systems, involve the vendor, your safety lead and qualified engineers (see modules 109 and 111).
Picture the week before a regulator's examination at a regional bank holding company with 6,000 employees. The examiners have sent a request list. They want the information security policy and evidence that the board approved it in the last 12 months. They want the standard for multifactor authentication and a report showing which systems do not meet it. They want the list of approved exceptions, who approved each one, and when each expires. They want to see that employees acknowledged the acceptable use policy, and that the procedure for removing a departing employee's access was followed for a sample of 25 people who left last quarter. Then they will ask the question that matters most: if your own documents say this is what you do, can you prove you did it?
That question is the subject of this lesson. A security program in a large organization runs on written rules: the policies the board approves, the standards that make them measurable, the procedures that tell staff what to do, and the exceptions that say where the rules are not yet met. Around those rules sits a compliance function that tracks which laws apply, maps them to controls, collects evidence, keeps a calendar of filings and certifications, and gives the board an honest answer about where the organization stands. When this machinery works, audits are routine and the board's annual certifications are true. When it does not, the gap between paper and practice becomes the evidence in an enforcement order, a False Claims Act complaint or a shareholder suit.
This lesson is written for compliance officers and the general counsel's office, CISOs and the people who write and maintain policy, internal auditors, executives and board members who approve policies and sign certifications, and the reporters, advocates and public officials who ask whether an organization's promises are real. The small-business version of this material, a short policy set and a simple training program, is in People, Policies and Training and Running the Program in module 73. The board's legal duties and the top-level information security policy are in Governance, Boards, Officers and Disclosure, and the frameworks themselves (CSF 2.0, SP 800-53, ISO 27001, CIS Controls) are explained in The Standards Map. This lesson does not repeat them. It covers how a large organization manages its rules and proves it follows them.
Inside this module, the security program's charter and organization are in lesson 1, metrics and board reports in lesson 3, SOC reports in lesson 5, certifications such as ISO 27001, PCI DSS validation and CMMC in lesson 6, internal audit and the three lines in lesson 8, privacy assessments in lesson 10, and the full program maturity checklist in lesson 14.
Why written rules matter in a large organization
In a 20-person office, the owner can tell everyone what to do. In an organization with thousands of employees, dozens of business units, hundreds of vendors and several regulators, the only way a rule reaches everyone, survives staff turnover and can be checked by an auditor is if it is written down, approved by someone with authority, published where people can find it, and tied to evidence. Written policy does four jobs at once:
- It states the organization's decisions. How much risk the board will accept, which data is sensitive, who may approve access, how fast critical vulnerabilities must be fixed. Without a written decision, each team makes its own.
- It assigns responsibility. A rule that does not name an owner has no owner.
- It is the yardstick for audits and examinations. Internal auditors, external auditors, SOC 2 auditors, PCI assessors and bank examiners begin by reading the policy and then test whether practice matches it. A policy that promises more than the organization does is worse than a modest policy that is followed, because the gap is a finding.
- It is evidence in enforcement and litigation. Regulators read policies to decide whether a failure was an isolated slip or a sign that the program existed only on paper, and they read them against what the company told customers and investors.
What the rules require in writing
Almost every regulatory regime that reaches a large organization requires written security policy, and several say who must approve it and how often:
| Rule | What it requires in writing | Approval and review |
|---|---|---|
| 23 NYCRR 500.3 (New York financial services, insurance and banking licensees) | A written policy or policies, based on the risk assessment, covering 15 listed areas from information security and data governance to vendor management, incident response and vulnerability management. | "Approved at least annually by a senior officer or the covered entity's senior governing body."1 |
| FTC Safeguards Rule, 16 CFR Part 314 (non-bank financial institutions, including auto dealers, mortgage brokers, tax preparers) | A written information security program with a written risk assessment that sets criteria for rating risks and for how each will be mitigated or accepted. | The Qualified Individual reports "in writing, regularly and at least annually" to the board on the program's status and compliance.2 |
| Interagency Guidelines, 12 CFR Part 30, Appendix B (national banks; parallel rules for other banks) | A written information security program. | The board or a board committee approves the program and oversees it; management reports to the board at least annually.15 |
| HIPAA Security Rule, 45 CFR 164.316 (health plans, providers, clearinghouses, business associates) | Written policies and procedures for every standard, and written records of required actions and assessments. | Kept six years from creation or the date last in effect, available to the people who carry them out, and reviewed periodically and updated as needed.16 |
| PCI DSS v4.0.1, Requirement 12 (anyone that stores, processes or transmits payment card data) | An overall information security policy, with roles and responsibilities, plus documented procedures for each requirement. | The policy is reviewed at least once every 12 months and updated as needed (the standard is licensed; described, not quoted).17 |
| NIST SP 800-53 Rev. 5 (federal agencies and many contractors; widely used elsewhere) | Every control family opens with a "Policy and Procedures" control (AC-1 for access control, IR-1 for incident response and so on). | Reviewed and updated at a frequency the organization defines, and after events it defines.18 |
| NIST CSF 2.0, Govern function, Policy category (voluntary; used by many state agencies, including Idaho's) | Cybersecurity risk management policy is established, communicated and enforced (GV.PO-01). | Reviewed, updated, communicated and enforced to reflect changes in requirements, threats, technology and mission (GV.PO-02).1920 |
Public companies have a related duty. Since December 2023, every annual report on Form 10-K must describe, under Item 1C, the company's processes for assessing, identifying and managing material cyber risks and how the board oversees them (Regulation S-K Item 106).21 Those descriptions are usually drafted from the company's own policies. If the 10-K says the company follows a framework and runs a risk-based program, the policy library is where an investor, an SEC examiner or a plaintiff will look to see whether that is true.
Policies as evidence of an effective program
Two federal documents set the standard that prosecutors and courts use to judge whether any compliance program, including a security and privacy compliance program, is real.
The first is Chapter 8 of the U.S. Sentencing Guidelines, which governs how federal courts sentence organizations. Section 8B2.1 describes an "effective compliance and ethics program." The organization must "exercise due diligence to prevent and detect criminal conduct" and "otherwise promote an organizational culture that encourages ethical conduct and a commitment to compliance with the law." The program must be "reasonably designed, implemented, and enforced so that the program is generally effective." The first of its seven minimum elements is that the organization "establish standards and procedures to prevent and detect criminal conduct."3 An organization that had such a program in place when an offense occurred can have its culpability score cut by three points under section 8C2.5(f), which can lower a fine substantially, unless it unreasonably delayed reporting the offense or high-level personnel took part in it.3
The second is the Justice Department Criminal Division's Evaluation of Corporate Compliance Programs, last updated in September 2024 and still the current version on Nov. 15, 2026. It tells prosecutors to ask three questions: "Is the corporation's compliance program well designed?" "Is the program being applied earnestly and in good faith?" and "Does the corporation's compliance program work in practice?"4 On policies, it asks what the company's "process for designing and implementing new policies and procedures" is, how the company "communicated its policies and procedures to all employees and relevant third parties," and whether the company "track[s] access to various policies and procedures" to learn which ones employees actually use.4 Those questions translate directly into the policy lifecycle described below.
Neither document is a cybersecurity rule. Both matter to security and privacy programs because a data breach that leads to false statements, a hidden incident, a false certification to the government or a sanctions violation can become a criminal or civil fraud matter, and because the same tests (is it designed, is it resourced, does it work) are now the way regulators of every kind talk about programs. The former Uber security chief's conviction for concealing a 2016 breach, described in 108 lesson 4, shows how a security incident becomes a criminal case.
The policy hierarchy: policy, standard, procedure, guideline
Large organizations that manage policy well keep four kinds of documents separate. The names vary (some call standards "requirements" or "control statements," some call procedures "runbooks" or "work instructions"), but the layers are the same.
| Layer | What it says | Typical length | Who approves | How often it changes | Example |
|---|---|---|---|---|---|
| Policy | What the organization requires and why; scope; who is responsible; authority of the documents below it; consequences of violation. | 2 to 10 pages each; a large organization may have one master policy and 10 to 20 topic policies. | Board or board committee for the master policy; executive risk committee or senior officer for topic policies. | Reviewed at least yearly; changed when strategy, law or risk appetite changes. | "Access to organization systems requires strong authentication proportionate to the sensitivity of the data." |
| Standard | The specific, measurable rule that carries out the policy. Mandatory. | A few pages to a few dozen; often a table of control statements. | CISO, chief privacy officer or a policy committee. | As technology and threats change; often several times a year. | "Multifactor authentication is required for all remote access, all administrative access and all email and cloud applications. SMS codes are not permitted for administrators." |
| Procedure | Who does what, in what order, with which tool, to meet the standard. Mandatory for the people who perform it. | Step lists, screenshots, checklists. | The manager of the team that performs it, with security review. | Whenever the tool or process changes. | "How the service desk verifies identity before resetting a token." |
| Guideline | Recommended practice. Not mandatory; explains how to meet a standard well. | Short articles, FAQs. | Any owning team. | Freely. | "Choosing a strong passphrase." |
Below all four sits the layer auditors care about most: records. Tickets, logs, configuration reports, signed attestations, training completions and approved exceptions are the proof that the rules above were followed.
The reasons for keeping the layers apart are practical. A policy approved by the board should not need to go back to the board because the help desk changed ticketing tools. A standard should be specific enough to test: an auditor can check "MFA on all remote access," but cannot test "appropriate access controls." Procedures written by the people who do the work are more accurate than procedures written by a policy office. And guidelines keep good advice from being turned into mandatory rules that generate thousands of exceptions.
A public sector example: the Idaho Technology Authority
Idaho's state government uses exactly this structure, and its documents are public. The Idaho Technology Authority (ITA), created under Idaho Code 67-831 and following, publishes enterprise IT documents for state agencies in three types marked by letter: P for policy, S for standard and G for guideline. The number shows the category: the 1000 series is general, and the 4000 and 4500 series policies cover security.22 The security policies include P4110 (Agency Cybersecurity Coordinator), P4130 (Information Systems Classification), P4140 (Cybersecurity Framework), P4502 (Privilege Access Management), P4520 (Patching and Vulnerability Management) and P4590 (Cybersecurity Incident and Breach Response Management and Reporting).22 P4140, revised Aug. 20, 2024, says agencies "shall adopt the current NIST Cybersecurity Framework as the basis for assessing their cybersecurity programs."20
How those documents are made is itself written down. ITA Policy P1015, established Aug. 5, 2024, sets the process: any of the ITA, its subcommittees, technical working groups or interested parties may propose a document; the proposer drafts it on ITA templates, with a redline for changes, and checks federal digital accessibility requirements; ITA staff must receive it at least four weeks before the committee meeting; staff send it out for a 14-day comment period; major comments trigger another round; and the responsible subcommittee votes. Policies approved by a subcommittee take effect on approval and are subject to review by the full ITA at its next regular meeting; standards and guidelines take effect on subcommittee approval and are reported to the ITA.23 P1015 does not set a fixed review cycle, which matters because most regulated private organizations must review at least yearly.
Exceptions have their own document. ITA Guideline G120 says an agency that cannot comply with an ITA policy or standard, or wants to buy or keep equipment or software that does not meet an enterprise standard, submits an Exemption Request Form and a letter from the agency director to the ITA chair. ITA staff review the request and make a recommendation; exemptions from policies go to the ITA, and exemptions from standards go to the IT Leadership Council, where an agency representative briefs the members. An exemption may be granted for a period "not to exceed twenty-four (24) months," and an agency that wants to extend one must submit a new request at least 90 days before it expires.5 G120's footer still shows an effective date of Feb. 19, 2003 and a revision of June 30, 2004, while its revision history lists an update on April 20, 2021, so readers should check the current form on the ITA site.
That model has what large organizations need: a typed hierarchy, a public drafting process with comment, a named approver for each type, and a time-limited, documented exception path. It also has what an auditor or reporter can check: exemption requests are public records under the Idaho Public Records Act (Idaho Code Title 74, Chapter 1), subject to its exemptions for security-sensitive information.
How many policies, and how to organize them
There is no correct number. The New York rule's 15 topics are a useful checklist for any organization: information security; data governance, classification and retention; asset inventory, device management and end of life management; access controls, including remote access and identity management; business continuity and disaster recovery; systems operations and availability; systems and network security and monitoring; security awareness and training; systems and application security and development; physical security and environmental controls; customer data privacy; vendor and third-party service provider management; risk assessment; incident response and notification; and vulnerability management.1
Large organizations usually choose one of two designs:
- One master policy plus topic policies. A short information security policy approved by the board (the sample in 108 lesson 4 is an example) delegates authority to 10 to 20 topic policies approved by an executive committee. This keeps the board's document short and stable.
- One consolidated policy organized by framework. A single policy with a section for each CSF 2.0 category or each control family, with standards attached. This is easier to map to one framework but harder to keep current, because every change reopens the whole document.
Either works if three things hold: every requirement appears in exactly one place (duplicates drift apart), every document has a named owner and a next-review date, and every mandatory statement can be tested.
The policy lifecycle: write, approve, publish, attest, review
A policy library is a living set of documents. The lifecycle below is common to well-run programs and matches what the Justice Department guidance, the New York rule and NIST ask about.
Drafting
Good policy drafting starts with three inputs: the risk assessment (what the organization is protecting against), the requirements register (which laws, contracts and standards apply), and current practice (what the organization actually does today). A draft that ignores current practice either describes a world that does not exist or creates hundreds of exceptions on the day it is approved.
Practical drafting rules that experienced policy teams follow:
- Use "must" for requirements and "should" only for recommendations, and define both in the policy. Auditors will treat "should" as optional.
- Write for the person who must comply. A policy for 6,000 employees should be readable by a branch manager or a nurse. Put technical detail in standards.
- Name roles, not people. "The CISO" survives a resignation; "Pat Smith" does not.
- Tie each requirement to its sources. A short citation (for example "NYDFS 500.12; PCI DSS 8.4; CSF PR.AA-03") in the standard, or a link in the governance, risk and compliance (GRC) tool, lets anyone see why a rule exists and which rules are affected if it changes.
- Do not copy paid standards. ISO/IEC 27001 and 27002, PCI DSS and the AICPA Trust Services Criteria are licensed. A policy can say it follows them and use its own words.
- Check that it can be met. Before approval, ask the teams that run the systems whether the standard is achievable and by when; build in a phased effective date if it is not.
Review and approval
The approval path should be written in the policy on policies (a sample appears later in this lesson). A common pattern:
- The document owner drafts or revises it and records what changed and why.
- Subject matter reviewers comment: IT operations, the affected business units, human resources (anything touching discipline, monitoring or employee data), the privacy office, the general counsel's office (for legal requirements and anything that becomes a promise to customers), internal audit (as an observer, to protect its independence), and labor relations where a union contract covers the subject.
- The policy committee or CISO approves standards and procedures; an executive risk committee approves topic policies; the board or its designated committee approves the master policy. In New York-regulated entities, the senior officer or senior governing body approves at least once a year.1
- Approval is recorded with the date, version number and approver, and minutes show what was discussed. A board that approves a 60-page policy package in two minutes without discussion has created a record, but not a good one.
Publishing, communicating and attesting
A policy that employees cannot find does not exist for compliance purposes. Large organizations publish policies on a single intranet site or policy portal, with the current version, effective date, owner and a plain-language summary, and keep superseded versions in an archive. Vendors and contractors who must follow a policy receive it under the contract.
Attestation means each person confirms, usually electronically, that they have read and will follow the policies that apply to them. It is typically collected at hire, each year and after a material change, and it is often combined with annual security awareness training. Attestation rates (for example, 98.7 percent of active employees attested within 30 days) are a standard compliance metric and are reported to leadership in many programs; see lesson 3.
Attestation proves that someone clicked. It does not prove understanding. The Justice Department's guidance asks the harder question: whether the company tracks which policies employees actually open and use, and whether it has tested employees' understanding.4 Programs that take this seriously use short quizzes, role-specific guidance (the developer gets the secure coding standard; the accounts payable clerk gets the payment verification procedure), and data from the policy portal and help desk to find policies that are confusing or ignored.
Yearly review and change triggers
Every document needs a next-review date. A yearly review is the norm for policies (required for NYDFS-covered entities and for the overall PCI DSS policy), while standards and procedures are often reviewed yearly or every two years. Reviews should also be triggered by events:
- a new or amended law or rule (for example, the FTC Safeguards Rule's notification requirement effective in 2024, or the CCPA regulations effective Jan. 1, 2026);
- a significant incident or near miss, or a finding from an audit, examination or penetration test;
- a merger, acquisition or divestiture (see lesson 11);
- a major technology change, such as a move to a new cloud provider or the adoption of generative AI tools (see lesson 13);
- a large number of exceptions against the same requirement, which usually means the requirement is wrong, unfunded or premature.
A review that changes nothing should still be recorded, with the reviewer's name and date. "Reviewed, no changes" is evidence; silence is not.
Version control, retention and retirement
Keep every approved version with its effective dates. Regulators and courts ask what the policy said on the date of an incident, not what it says today. HIPAA requires written policies and records to be kept six years from creation or from the date they were last in effect, whichever is later.16 New York requires covered entities to keep the records supporting their annual compliance filing for five years.9 When a policy is retired or merged into another, record where its requirements went. When a policy is in litigation or under investigation, a legal hold overrides normal retention; see module 73 lesson 17.
Exceptions and risk acceptance
No large organization meets every standard on every system all the time. A hospital has an imaging device whose vendor supports only an old operating system. A manufacturer has a controller that cannot take an agent. A bank acquires a company that uses a different identity provider. A county has a legacy tax system that cannot do multifactor authentication until it is replaced in 18 months. The program's honesty depends on how those gaps are handled.
Three terms that get confused
- Exception (or waiver, or exemption)
- Formal, time-limited permission not to meet a specific requirement for a specific system, process or group, granted through a documented process.
- Risk acceptance
- A decision by someone with the authority to do so that the remaining risk is within the organization's appetite and will be carried rather than reduced further. Every exception involves a risk acceptance, but not every risk acceptance is an exception: a risk can be accepted where no rule applies.
- Compensating control
- A different control that reduces the same risk when the required control cannot be used, for example isolating an unsupported device on its own network segment with strict allowed connections. PCI DSS has a formal compensating controls method with a worksheet, and since version 4.0 a "customized approach" in which an entity designs its own control to meet a requirement's stated objective, with the assessor testing it.17
HIPAA has a related idea built into the rule. Its "addressable" implementation specifications are not optional: a covered entity must implement each one if it is reasonable and appropriate, and if not, must document why and implement an equivalent alternative measure if that is reasonable and appropriate.16 The documentation is the exception record. HHS's proposed Security Rule update, published Jan. 6, 2025, would remove the distinction between required and addressable specifications, with limited exceptions; on Nov. 15, 2026 it remained a proposal, listed by HHS among long-term actions.24 See 109 lesson 3.
Who may accept risk
The most important rule in an exception process is that the person who wants the exception is not the person who approves it, and that the approver's level matches the size of the risk. A common pattern, written into the exception procedure and tied to the organization's risk rating scale from 108 lesson 2:
| Residual risk rating | Who may approve | Maximum term (example) | Reported to |
|---|---|---|---|
| Low | CISO's delegate (for example, the GRC manager) with the system owner | 12 months | Quarterly exception report |
| Medium | CISO and the business unit leader who owns the system | 12 months | Quarterly exception report to the executive risk committee |
| High | Executive risk committee or the senior executive who owns the business process (for example the COO), after CISO review | 6 months, renewable once | Board risk or audit committee at the next meeting |
| Critical, or a requirement of law or contract | Generally not available as an exception: a gap in a legal requirement is a compliance failure to be fixed and, where required, reported. A temporary plan approved by the CEO and reported to the board may be needed. | Only as long as the remediation plan requires | Board; regulators where the rule requires |
The last row matters. An internal exception cannot waive a law. A New York-regulated company that has not met a Part 500 requirement must say so in its April 15 filing with a remediation timeline, rather than certifying compliance;9 a defense contractor whose SPRS score or plan of action does not reflect reality has a False Claims Act problem, not an exception (see the cases later in this lesson).
The exception register
Every exception goes into a single register (in a GRC tool or, in smaller programs, a controlled spreadsheet) with at least: the requirement not met; the system, process or group affected; the business reason; the risk rating before and after compensating controls; the compensating controls and who maintains them; the approver and approval date; the expiry date; the remediation plan and its owner; and links to evidence. The register is reviewed at least quarterly, and leadership sees four numbers: open exceptions, exceptions by risk rating, exceptions past their expiry date, and exceptions renewed more than once.
Two patterns signal trouble. The first is the permanent exception, renewed every year for a decade. The second is the exception that never existed: a gap everyone knows about that was never recorded, so leadership believes the standard is met. Internal audit should sample both, and lesson 8 explains how.
When an exception's remediation requires a change to production systems, such as replacing an unsupported server or turning on multifactor authentication for a legacy application, plan the change, test it, and have a way back if it fails.

The compliance function and its independence
Security and compliance are related but different jobs. The security function designs and runs controls to reduce risk. The compliance function identifies which laws, regulations, contracts and standards apply, translates them into requirements, checks whether the organization meets them, and reports honestly when it does not. In the language of the three lines model described in lesson 8, business units and IT operations are the first line, the CISO's governance team, risk management, compliance and privacy are usually second-line functions that set rules and monitor, and internal audit is the third line that gives independent assurance.
What a cyber compliance team does
- Keeps the requirements register: every law, rule, contract clause and certification that imposes a security or privacy obligation, with the business units and systems it reaches. In a national company this includes the federal sector rules, more than a dozen state privacy laws, all 50 state breach laws (see 108 lesson 11), foreign laws where it operates, and hundreds of customer contracts with security schedules.
- Watches for change: proposed and final rules, enforcement actions against peers, new guidance, and contract renewals.
- Maps requirements to controls and owns the control library (described in the next section).
- Runs the policy process or supports the CISO's office in running it.
- Plans and supports audits, examinations and assessments, and coordinates evidence so control owners are not asked the same question five times.
- Tests controls as a second-line check before auditors do, and tracks findings to closure.
- Prepares regulatory filings and certifications and the evidence behind them.
- Reports the organization's compliance position to executives and the board.
Independence and reporting lines
Where compliance sits, and whom it can go to, determines whether it can say no. The Justice Department guidance asks "Where within the company is the compliance function housed," whether compliance has "sufficient autonomy from management, such as direct access to the board of directors or the board's audit committee," whether compliance and control personnel have "sufficient direct or indirect access to relevant sources of data," and how "the assets, resources, and technology available to compliance and risk management compare" to those used elsewhere in the company.4 The Sentencing Guidelines require that the board be "knowledgeable about the content and operation" of the program, that high-level personnel oversee it, and that the people with day-to-day responsibility have adequate resources, appropriate authority and direct access to the board or a board committee.3
For cybersecurity compliance, the reporting choices look like this:
| Where cyber compliance sits | Strengths | Risks |
|---|---|---|
| Inside the CISO's organization (a GRC team) | Close to the controls; understands the technology; fast. | Marks its own homework; may soften findings about the CISO's own program. |
| In an enterprise compliance or risk function under a chief compliance officer or chief risk officer | Independent of the CISO; sees cyber next to other compliance risks; direct line to the board. | May lack technical depth; can drift into checklist reviews. |
| Split: technical GRC in the CISO's office, regulatory compliance oversight in enterprise compliance | Common in large banks and insurers; combines depth and independence. | Needs a clear written division of duties, or work is duplicated and gaps open. |
Whatever the structure, three protections matter: the head of compliance can reach the board committee without management's permission, compliance has read access to the data it needs (identity systems, vulnerability scans, ticketing), and compliance staff cannot be disciplined for honest findings. The same logic is why many regulators want the CISO to report to the board directly, discussed in lesson 1.
The seven elements applied to cyber
The Sentencing Guidelines' seven minimum elements were written for all compliance, but they translate directly into a security and privacy program. The left column follows section 8B2.1(b); the right column is this lesson's application.3
| Element (USSG 8B2.1(b)) | What it looks like in a security and privacy program |
|---|---|
| 1. Standards and procedures to prevent and detect criminal conduct | The policy hierarchy in this lesson, including rules against concealing incidents and against false statements in certifications and customer questionnaires. |
| 2. Board knowledge and oversight; high-level personnel responsible; day-to-day responsibility delegated with resources, authority and access to the board | Board committee charter covering cyber; a named executive owner; a CISO and compliance lead with budget and a direct line to the committee (108 lesson 4). |
| 3. Reasonable efforts not to give substantial authority to people with a history of illegal or inconsistent conduct | Background checks for privileged administrators and security staff, consistent with law; reviews of who holds the most powerful access. |
| 4. Communication and practical training | Policy publication, attestation and role-based training (108 lesson 9). |
| 5. Monitoring, auditing, periodic evaluation, and a reporting system that allows anonymous or confidential reports without fear of retaliation | Control monitoring, internal audit, and a hotline that accepts security concerns, not just financial ones. |
| 6. Consistent promotion and enforcement through incentives and discipline | Security objectives in executive scorecards; the same consequences for a senior executive who bypasses a control as for a junior employee. |
| 7. Appropriate response to detected conduct and changes to prevent recurrence | Incident reviews that change policy and controls; corrections to filings and certifications when errors are found. |
Section 8B2.1(c) adds that the organization "shall periodically assess the risk of criminal conduct" and adjust each element accordingly.3 For a security program, the cyber risk assessment and the compliance risk assessment should feed each other.
Compliance in public agencies
State and local governments rarely have a chief compliance officer for cybersecurity. The functions still exist: a state CIO or CISO sets policy (in Idaho, through ITA documents and the Office of Information Technology Services), agencies attest to compliance, and independent checks come from legislative auditors, federal program auditors and grant conditions. Federal programs bring their own written requirements: the IRS Publication 1075 safeguards for federal tax information, the FBI's Criminal Justice Information Services (CJIS) Security Policy for law enforcement data, and the Centers for Medicare and Medicaid Services rules for state Medicaid systems each require written policies, assessments and reports. A county or state agency that touches all three keeps a requirements register just as a bank does. Module 73 lesson 16 covers these at small scale.
One control, many rules: mapping and unified control frameworks
A large organization may be subject to a dozen overlapping security requirements: NYDFS Part 500, the FTC Safeguards Rule or banking guidelines, HIPAA for its health plan, PCI DSS for card payments, SOC 2 commitments to customers, ISO 27001 certification for European clients, NIST SP 800-171 and CMMC for a defense contract, and state privacy laws. Most of them ask for the same things in different words: know your assets, control access, use multifactor authentication, encrypt sensitive data, log and monitor, train people, manage vendors, respond to incidents, test. Running a separate compliance program for each one wastes money and produces inconsistent answers.
The solution is a single control library (sometimes called a unified control framework or common control framework): one set of control statements, each with an owner, a test procedure and evidence, mapped to every requirement it satisfies. Test the control once; report the result against every mapped rule.
NIST informative references and OLIR
NIST supports mapping in three ways. The CSF 2.0 core lists "informative references" for each outcome, linking it to sections of other standards and guidance.19 The National Online Informative References (OLIR) program publishes mappings submitted by NIST and outside organizations between NIST documents and other standards, regulations and frameworks, with tools to view and compare them, and the CSF 2.0 Reference Tool lets users browse and export the core with its references.25 In 2026 NIST published SP 1347, a quick-start guide that explains what informative references are, how they support CSF 2.0 outcomes, and how to get them by direct download, through the CSF 2.0 Reference Tool or through OLIR, with two sample use cases.7
NIST IR 8477, Mapping Relationships Between Documentary Standards, Regulations, Frameworks, and Guidelines, published in final form Feb. 26, 2024 (the same day as CSF 2.0), sets out how to describe a mapping carefully.6 It describes three styles: a simple concept crosswalk (these two items are related), a set theory relationship mapping that says how two items relate (one is a subset of the other, they are equal, they intersect, one is a superset, or they are not related), and a supportive relationship mapping that says how one item supports another. The set theory style matters to compliance teams, because "intersects with" is the honest description of most mappings between a regulation and a framework: the two overlap, but neither fully covers the other.
The Secure Controls Framework and commercial control libraries
The Secure Controls Framework (SCF) is a free control library maintained by the SCF Council, a private group. It organizes controls into 34 domains as of release 2026.2 (dated July 7, 2026, which added a quantum security domain) and publishes mappings to a long list of laws, regulations and frameworks, using the set theory relationship method from NIST IR 8477 for its published mappings.8 Commercial GRC vendors and the large audit firms sell their own control libraries with mappings, usually as part of a GRC platform. The Cloud Security Alliance's Cloud Controls Matrix and the Shared Assessments program's Standardized Information Gathering (SIG) questionnaire are other widely used libraries, the second focused on vendor risk questionnaires. Which library an organization uses matters less than three practices: keep one library, record the source and date of each mapping, and have someone who knows the rule check the mappings for the regulations that carry the biggest penalties.
A mapping is not proof
Mappings are where compliance programs get into trouble, for three reasons:
- The words differ. NYDFS 500.12, as amended, requires multifactor authentication for any individual accessing any of the covered entity's information systems, with limited exceptions approved by the CISO;1 PCI DSS requires it for access into the cardholder data environment; NIST SP 800-171 Rev. 2 requires it for privileged accounts and for network access to non-privileged accounts. One control can satisfy all three only if it is written to the strictest one.
- Scopes differ. PCI applies to the cardholder data environment, CMMC to the systems that handle federal contract or controlled unclassified information, HIPAA to electronic protected health information. A control that is fully implemented in one scope can be missing in another.
- Evidence differs. A SOC 2 auditor samples over a period; a PCI assessor tests on the assessment date; a CMMC assessor scores each requirement as met or not met. A single test may need to produce evidence in several forms.
The defense cases later in this lesson show what happens when a self-assessment score is built on a mapping rather than on the actual systems in scope: Georgia Tech's score of 98 was alleged to rest on a "fictitious" or "virtual" campus-wide environment that did not exist.14 108 lesson 3 covers the frameworks and the hazards of framework claims in more detail.
GRC platforms in plain terms
"GRC" stands for governance, risk and compliance. A GRC platform is software that keeps the program's records in one database instead of hundreds of spreadsheets and email threads. The products are sold by large enterprise software companies, by security specialists and by newer firms focused on automating SOC 2 and ISO 27001 evidence. Most offer some combination of these modules:
| Module | What it holds | Why it helps |
|---|---|---|
| Policy management | Policies, standards and procedures with owners, versions, approval workflows, review dates and attestation campaigns. | Proves who approved what and when; sends review reminders; tracks attestation rates. |
| Requirements and control library | Laws, rules and standards, the control statements, and the mappings between them. | Shows which rules a failed control affects. |
| Risk register | Risks, ratings, owners, treatment plans and risk acceptances. | Ties exceptions to risk decisions; feeds board reports. |
| Evidence and testing | Test procedures, test results, uploaded or automatically collected evidence. | One evidence request serves several audits. |
| Issues and remediation | Audit findings, exam findings, test failures, with owners and due dates. | Tracks closure; flags overdue items. |
| Third-party risk | Vendor inventory, questionnaires, SOC reports and contract terms. | Connects vendor risk to the same risk scale (108 lesson 8). |
| Audit management | Audit plans, workpapers and requests. | Used by internal audit, sometimes separately from the second line to keep independence. |
Buying and running a GRC platform
The common mistakes are predictable. Organizations buy a large platform before they have a control library, then spend a year configuring workflows around an undefined process. They let the tool's default framework become their policy. They automate evidence collection for controls that are poorly designed, producing well-organized proof of the wrong thing. And they give the platform to one administrator who leaves. A sensible order is: write the policy on policies and the exception procedure; build the requirements register and control library (a spreadsheet is fine at first); pilot with one regulation and one audit; then buy software that fits the process. Ask any vendor for its integration list, how mappings are updated when rules change and who checks them, how data can be exported if the contract ends, and whether the platform itself has a current SOC 2 Type 2 report (lesson 5).

Evidence and continuous control monitoring
What counts as evidence
Auditors and examiners distinguish two questions. Design: if the control works as described, would it meet the requirement? Operating effectiveness: did it actually work, consistently, over the period? A policy and a configuration screenshot can show design. Operating effectiveness takes evidence over time: for a quarterly access review, the four review records with reviewer names, dates and the changes made; for a joiner-mover-leaver process, a population of all departures in the period and a sample showing access removed on time.
Good evidence is complete (the whole population, not a hand-picked set), accurate (system-generated where possible, with the query or report parameters shown), dated, attributable to a person or system, and retained under the records schedule. A common audit test is to ask for the full population of an event (all new privileged accounts this year), confirm its completeness against another source, then select a sample and trace each item to its approval. If the organization cannot produce a complete population, the auditor cannot rely on the control.
Continuous control monitoring
Continuous control monitoring (CCM) means testing controls automatically and frequently, often daily, instead of once a year. Typical automated tests in a large organization: the share of accounts with multifactor authentication enabled; administrator accounts created outside the approval workflow; endpoints missing the detection agent; cloud storage open to the internet; critical vulnerabilities past their due date (for example, the dates CISA sets for its Known Exploited Vulnerabilities catalog); backups that did not complete; departed employees whose accounts are still active. Results flow into the GRC platform and dashboards.
CCM has two benefits: problems are found in days, not at the annual audit, and the same data becomes evidence. Its limits are just as real. Automated tests check what is easy to measure, which is not always what matters; they depend on accurate asset inventories (an endpoint that is not in the inventory is not tested); and a green dashboard can create false comfort. The New York rule now requires covered entities to maintain an asset inventory with owner, location, classification, support expiration date and recovery time objectives, which is the foundation any automated monitoring needs.1 Keep periodic manual testing and internal audit sampling alongside automation.
The compliance calendar for a large organization
A large regulated organization lives by a calendar of fixed filing dates, recurring reviews and assessment windows. The compliance team owns the master calendar, assigns each item an owner and a lead time, and works backward from each deadline to when evidence must be collected. The table lists real deadlines and frequencies; the infographic after it places them on a year for an illustrative company whose fiscal year ends Dec. 31.
| Obligation | Rule and version | Deadline or frequency | Who signs or owns | Evidence behind it |
|---|---|---|---|---|
| Annual report cyber disclosure (Item 1C) | Regulation S-K Item 106; Form 10-K (SEC Release 33-11216, 2023) | With the 10-K: 60 days after fiscal year end for large accelerated filers, 75 days for accelerated filers, 90 days for others | Disclosure committee; CEO and CFO certify the 10-K | Policies, risk assessment process, board and committee minutes, CISO qualifications2126 |
| Material cyber incident report | Form 8-K Item 1.05 | Four business days after the company determines an incident is material | Disclosure committee | Materiality decision record (108 lesson 4)21 |
| Annual certification or acknowledgment of noncompliance | 23 NYCRR 500.17(b), as amended Nov. 1, 2023 | April 15 each year, covering the prior calendar year | Highest-ranking executive and the CISO | Records supporting the filing, kept five years9 |
| Policy approval; risk assessment update; penetration test; CISO report to board | 23 NYCRR 500.3, 500.9, 500.5, 500.4 | Each at least annually | Senior officer or board; CISO | Approvals, assessment, test report, board report127 |
| Written report to the board | FTC Safeguards Rule, 16 CFR 314.4(i) | Regularly and at least annually | Qualified Individual | Report on program status and compliance2 |
| Penetration test; vulnerability assessments | 16 CFR 314.4(d)(2) | Penetration test yearly; vulnerability assessments at least every six months and after material changes | Qualified Individual | Test and scan reports2 |
| Notice to the FTC of a notification event | 16 CFR 314.4(j), effective May 13, 2024 | No later than 30 days after discovery, if at least 500 consumers' information is involved | Qualified Individual | Incident record2 |
| Report to the board on the information security program | Interagency Guidelines, 12 CFR Part 30, Appendix B | At least annually | Management | Board report15 |
| Risk analysis and evaluation; policy review | HIPAA, 45 CFR 164.308(a)(1), 164.308(a)(8), 164.316(b) | No fixed interval in the current rule: "periodic" and in response to change; records kept six years. The 2025 proposal would add yearly reviews. | Security official | Risk analysis, evaluation report1624 |
| Payment card assessment | PCI DSS v4.0.1; validation level set by the card brands and acquiring bank | Yearly Report on Compliance or Self-Assessment Questionnaire with Attestation of Compliance; quarterly external vulnerability scans; policy review every 12 months | Executive signer of the AOC; QSA where required | ROC or SAQ, scan reports (lesson 6)17 |
| SOC 2 Type 2 report | AICPA attestation standards (lesson 5) | A review period chosen by the company, commonly 6 or 12 months, with a report after the period ends | Management assertion | Control evidence across the period |
| CMMC affirmation | 32 CFR 170.22 (effective Dec. 16, 2024) | After each assessment and plan of action closeout, and annually thereafter, entered in SPRS | Affirming Official, a senior representative of the contractor | Assessment results and continuing compliance10 |
| Privacy risk assessments and cybersecurity audits (California) | CPPA regulations, approved Sept. 23, 2025, effective Jan. 1, 2026 | Risk assessments for processing already under way by Dec. 31, 2027; information about 2026 and 2027 assessments submitted by April 1, 2028; first cybersecurity audit certification by April 1, 2028 for businesses with 2026 revenue over $100 million, 2029 for $50 to $100 million, 2030 for under $50 million | A business executive signs submissions | Assessments and audit reports (lesson 10)28 |
| Regulation S-P incident response program (broker-dealers, investment advisers, funds, transfer agents) | SEC amendments adopted May 16, 2024 | Compliance from Dec. 3, 2025 (larger entities) and June 3, 2026 (smaller entities); customer notice within 30 days of a breach | Chief compliance officer | Written policies and procedures29 |
| Exemptions from statewide IT policy (Idaho agencies) | ITA Guideline G120 | Up to 24 months; renewal request at least 90 days before expiry | Agency director | Exemption request and ITA decision5 |
Three more things belong on the calendar even though no rule sets the date: the yearly review of every policy and standard, the quarterly exception register review, and the insurance renewal, whose application asks many of the same questions as an audit (lesson 9). Some organizations add the anniversary of each major customer contract's audit right. Federal financial regulators' examinations follow their own cycles and arrive with request letters; the FFIEC IT Examination Handbook booklets describe what examiners review.
The calendar for defense contractors changed during 2026. CMMC Phase I, which began Nov. 10, 2025, made self-assessments and annual affirmations a condition of many new contracts, and Phase II was to require third-party Level 2 assessments from Nov. 10, 2026. On July 13, 2026, the Department of War suspended Phase II's third-party and government-led assessment mandates in two memoranda (government-led assessments remain possible under Phase I authority), set up a CMMC Reform Task Force with a 60-day review, and opened a request for information that closed Aug. 14, 2026. The suspension did not withdraw the rule: DFARS 252.204-7012, NIST SP 800-171 Rev. 2, SPRS score postings, self-assessments and annual affirmations still apply where contracts require them.30 This lesson could not confirm the task force's final recommendations or a restart date as of Nov. 15, 2026; lesson 6 tracks the status.
Building and running the calendar
- Start from the requirements register: every rule with a date or frequency becomes a calendar entry with an owner.
- Work backward. A 10-K filed 60 days after year end needs the Item 1C draft reviewed by the disclosure committee weeks before; an April 15 New York filing needs a sub-certification process from control owners in February and March.
- Use sub-certifications. Many large organizations ask each control owner and business unit leader to certify, in writing, that their controls operated as described and to list exceptions, before the CEO and CISO sign. This mirrors the sub-certification practice many public companies use under Sarbanes-Oxley (lesson 8).
- Reconcile across filings. The 10-K, the New York certification, the SOC 2 management assertion, the PCI AOC, the CMMC affirmation and the cyber insurance application all describe the same program. Inconsistencies among them are easy for a regulator or plaintiff to find.
- Report calendar status to the executive risk committee monthly and to the board committee quarterly, with anything at risk of being late.
How prosecutors and sentencing courts judge a program
The Sentencing Guidelines and the Justice Department's compliance guidance were introduced above as yardsticks. For a large organization they also have direct consequences: they affect whether a company is charged, what kind of resolution it receives, whether it must hire a monitor, and the size of a fine.
What changed in 2025 and 2026
The Justice Department's corporate enforcement policies moved several times between 2024 and 2026. The Criminal Division's compliance evaluation guidance itself was last updated in September 2024, when it added questions about the management of emerging technologies such as artificial intelligence, about whistleblower protection and about the use of personal devices and messaging apps; it remained the current version on Nov. 15, 2026.431 Around it:
- August 2024: a temporary amendment to the Criminal Division's voluntary self-disclosure policy let a company that receives an internal whistleblower report still qualify for a presumption of declination if it reports to the department within 120 days, and the Criminal Division started a Corporate Whistleblower Awards Pilot Program.31
- May 2025: the Criminal Division issued a White-Collar Enforcement Plan, revised its corporate enforcement and self-disclosure policy and its monitor selection memorandum, and updated the whistleblower pilot to match the department's priorities. A crediting policy followed in June 2025.31
- March 10, 2026: the Deputy Attorney General issued the first department-wide Corporate Enforcement and Voluntary Self-Disclosure Policy, which largely extends the Criminal Division's 2025 approach to all criminal matters. How it will interact with the compliance evaluation guidance and other components' frameworks was, according to a law firm review published March 12, 2026, not yet clear.3132
The practical meaning for a security and privacy program has not changed: the department rewards companies that find problems, report them, cooperate and fix the root cause, and it judges compliance programs by evidence of operation, not by documents alone.
Applying the guidance to a security program
Questions a compliance officer or CISO can expect, adapted from the department's guidance to security subjects:4
- How are new security and privacy policies designed, and who is consulted? Are they based on a current risk assessment?
- Can employees find the policies that apply to them? Does the company know which policies are read and which are ignored?
- Do compliance and security staff have access to the data they need, such as identity, logging and vulnerability systems, to test whether controls work?
- How do the resources for security compliance compare with the resources for revenue-producing technology projects?
- When a control failed or an incident occurred, did the company identify the root cause and change the program? Are those "lessons learned" written down?
- For AI and other new technology, "Do controls exist to ensure that the technology is used only for its intended purposes?"
- "Does the company have an anti-retaliation policy?" Have people who raised security concerns been treated fairly?

Whistleblowers, the False Claims Act and cyber fraud
For a security program, whistleblower law is not an abstract topic. The people best placed to know whether a company's security certifications are true are its own security staff, IT administrators and compliance analysts. Federal law pays some of them for reporting, protects them from retaliation, and has produced most of the government's cybersecurity fraud cases.
The SEC whistleblower program
Under the Dodd-Frank Act, the SEC pays awards of 10 to 30 percent of the money collected when original information from a whistleblower leads to an enforcement action with sanctions of more than $1 million. In fiscal year 2025 the SEC received about 27,000 tips, awarded more than $60 million to 48 individual whistleblowers, and paid $170,221,335.88 from its Investor Protection Fund to whistleblowers (payments in a year can include awards made earlier).33 Information about a public company misstating its cyber risk management, hiding a material incident or keeping inadequate controls over data covered by Regulation S-P can be a tip like any other. SEC rules also prohibit any person from taking action to impede someone from communicating directly with the SEC about a possible violation, which is why confidentiality clauses in employment agreements, severance agreements and incident response engagement letters should say plainly that they do not limit reports to government agencies.
The False Claims Act and qui tam suits
The False Claims Act makes anyone who knowingly submits a false claim for payment to the federal government, or makes a false statement material to such a claim, liable for three times the government's damages plus a civil penalty per claim that is adjusted each year for inflation. Private persons, called relators, may file suit on the government's behalf under seal (a "qui tam" suit). If the government intervenes, the relator receives "at least 15 percent but not more than 25 percent" of the proceeds; if it declines and the relator wins alone, 25 to 30 percent. Employees who are fired, demoted, harassed or otherwise retaliated against for protected acts can recover reinstatement, "2 times the amount of back pay," interest and other damages, with a three-year deadline to sue.34
For cybersecurity, the claim is usually that a contractor or grant recipient represented that it met contract security requirements, such as DFARS 252.204-7012 and NIST SP 800-171, FAR 52.204-21, a FedRAMP requirement or a health data security clause, when it did not, and was paid anyway. The representation can be in a proposal, an SPRS score, a CMMC affirmation, an annual certification or a system security plan.
The Civil Cyber-Fraud Initiative and its cases
Deputy Attorney General Lisa O. Monaco announced the Civil Cyber-Fraud Initiative on Oct. 6, 2021, to use the False Claims Act against those who knowingly provide "deficient cybersecurity products or services," knowingly misrepresent "cybersecurity practices or protocols," or knowingly violate obligations "to monitor and report cybersecurity incidents and breaches." The announcement pointed to the act's whistleblower provision and its protection from retaliation.12 The initiative continued under the next administration: settlements were announced in 2025 and 2026, and the Assistant Attorney General for the Civil Division said in September 2025 that the department "will continue to pursue and litigate violations of cybersecurity requirements."14 Selected cases, with the amounts and allegations as the department or settlement reviews described them (settlements are not admissions of liability):
| Date | Organization | Amount | What was alleged | Who brought it |
|---|---|---|---|---|
| July 8, 2022 | Aerojet Rocketdyne | $9 million | Misrepresented compliance with cybersecurity requirements in federal contracts. Settled on the second day of trial. | Brian Markus, a former employee; his share was $2.61 million.35 |
| Oct. 22, 2024 | Pennsylvania State University | $1.25 million | Did not implement required controls in 15 DoD and NASA contracts or subcontracts from 2018 to 2023; reported DoD assessment scores with inaccurate dates for when missing controls would be in place; on some contracts did not use a cloud provider meeting DoD requirements. | Matthew Decker, former CIO of Penn State's Applied Research Laboratory; his share was $250,000.13 |
| Feb. 18, 2025 | Health Net Federal Services and Centene | $11,253,400 | False annual cybersecurity certifications on a TRICARE health contract; vulnerability scanning not done; audit warnings ignored. | As described in a settlement tracker36 |
| March 26, 2025 | MORSECORP | $4.6 million | Used an email host that did not meet FedRAMP Moderate; had not implemented all NIST SP 800-171 controls; reported an SPRS score of 104 when a consultant later scored it at -142. | Relator; company admitted facts36 |
| May 1, 2025 | Raytheon, RTX and Nightwing | $8.4 million | Used an internal system that did not meet DFARS and FAR requirements to handle covered defense and federal contract information. | As described in a settlement tracker36 |
| July 31, 2025 | Illumina | $9.8 million | Sold genomic sequencing systems with cybersecurity vulnerabilities to federal agencies, with an inadequate product security program and inaccurate representations about meeting ISO and NIST standards. | As described in a settlement tracker36 |
| Sept. 30, 2025 | Georgia Tech Research Corporation | $875,000 | A cyber-defense research lab ran without proper anti-malware protection until December 2021 and without a system security plan until at least February 2020; a December 2020 SPRS score of 98 covered a "fictitious" or "virtual" campus-wide environment. | Christopher Craig and Kyle Koza, former members of Georgia Tech's cybersecurity team; their share was $201,250.14 |
| Sept. 1, 2026 | Honeywell Aerospace | $2,042,518 | A business unit did not meet NIST SP 800-171 controls on one network from April 2020 through December 2023, as its DoD contract required. | Rachel Tenney, a former employee, filed in 2022; the government intervened in part Aug. 18, 2026; her share was $375,823.37 |
A settlement tracker that follows these cases counted 18 settlements under the initiative through Sept. 1, 2026, totaling at least $84,271,106; that total comes from a private tracker, not the department.36
What the cases teach compliance programs
- The scores and dates are the claims. Penn State, Georgia Tech and MORSECORP were about numbers and dates reported to the government. A compliance team that reviews every SPRS submission, plan of action date and annual certification before it is filed, and keeps the evidence, removes the most common basis for these cases.
- The relators were insiders in security and IT roles: a research lab's CIO, members of a cybersecurity team, employees. When they raised concerns internally first, the response became part of the story. A program that takes internal reports seriously, investigates them and corrects filings is both the right answer and the best defense.
- Correcting is cheaper than defending. The department's policies give credit for self-disclosure, cooperation and remediation. A compliance officer who finds that a past certification was wrong should have a written path to correct it, involving the general counsel's office, rather than leaving it in place.
- Scope matters. Georgia Tech's alleged problem was a score for an environment that did not exist; Honeywell's was one network. Document exactly which systems each certification covers.
Internal reporting channels that work
A hotline that accepts only accounting complaints, or routes security concerns back to the CISO whose program is the subject, does not meet the Sentencing Guidelines' fifth element. A workable design: reports can be made anonymously through a third-party service; the compliance or ethics office (not the security team) receives and logs them; security concerns are triaged by someone independent of the program being criticized; the board committee sees a summary of security-related reports and their outcomes; and the anti-retaliation policy names security and data protection concerns explicitly. The Justice Department's 120-day window for self-reporting after an internal whistleblower report, introduced in August 2024, is a reason to move quickly once a credible report arrives.31
Sample: Policy Management Procedure
This sample is a "policy on policies." It assumes a master information security policy approved by the board, such as the sample in 108 lesson 4, and sets how every document below it is written, approved, published, reviewed and retired.
Sample only. Adapt it to your organization's risks, systems and laws, have qualified people review it, and do not treat it as a substitute for a risk assessment.
Follows: NIST Cybersecurity Framework 2.0, GV.PO-01 and GV.PO-02 (Feb. 26, 2024); NIST SP 800-53 Rev. 5, the "-1" Policy and Procedures controls; 23 NYCRR 500.3 (as amended Nov. 1, 2023); 45 CFR 164.316(b); U.S. Sentencing Guidelines section 8B2.1(b)(1) and (b)(4); Justice Department Evaluation of Corporate Compliance Programs (Sept. 2024); Idaho Technology Authority P1015 (Aug. 5, 2024).
1. Purpose and scope
This procedure governs how [Organization name] creates, approves, publishes, communicates, reviews and retires information security and privacy policies, standards, procedures and guidelines ("governance documents"). It applies to every governance document issued under the Information Security Policy approved by the [Board / Risk Committee] on [date], in every business unit and subsidiary, and to vendors required by contract to follow them.
2. Document types
- Policy: states what [Organization name] requires and why, assigns responsibility and sets consequences. Mandatory.
- Standard: sets specific, measurable requirements that carry out a policy. Mandatory.
- Procedure: describes the steps a team follows to meet a standard. Mandatory for the people who perform it.
- Guideline: recommends good practice. Not mandatory.
In governance documents, "must" and "must not" state requirements; "should" states a recommendation. Each requirement appears in only one governance document; other documents refer to it.
3. Ownership
Each governance document has one owner, named by role, who keeps it current and accurate, answers questions about it, and proposes changes. The [Chief Information Security Officer] owns the Information Security Policy and assigns owners for other documents. The [GRC Manager] maintains the governance document register, which lists every document with its type, owner, version, approval date, effective date, next review date and the requirements it maps to.
4. Drafting and consultation
Owners draft on the approved template. Each draft records what changed, why, and which laws, regulations, contracts and frameworks it addresses. Before approval, the owner consults: the teams that must carry it out; [IT Operations]; [Human Resources] for anything affecting employee conduct, monitoring or discipline; the [Privacy Office]; the [general counsel's office] for legal requirements and external commitments; [Labor Relations] where a collective bargaining agreement applies; and [Accessibility] to confirm the published document meets the organization's accessibility standard. [Internal Audit] receives drafts for information only. The owner confirms with system owners that each new or stricter requirement can be met by the effective date, or proposes a phased effective date.
5. Approval authority
| Document | Approved by | Recorded in |
|---|---|---|
| Information Security Policy | [Board / Risk Committee], on recommendation of the [CEO] | Board or committee minutes |
| Topic policies | [Executive Risk Committee] | Committee minutes |
| Standards | [CISO] (security) or [Chief Privacy Officer] (privacy), after [Policy Review Group] review | Governance document register |
| Procedures and guidelines | Owning manager, with [Security GRC] review | Governance document register |
Approval records state the version, date, approver and a summary of discussion. Where a regulation requires approval by a senior officer or the board at a set frequency, the [GRC Manager] schedules it on the compliance calendar.
6. Publication and communication
Approved documents are published on the [policy portal] within [5] business days, with a plain-language summary, the effective date and the owner's role. Superseded versions are archived and remain retrievable. Material changes are announced to affected staff. Vendors bound by a governance document receive it through [Procurement / Vendor Management].
7. Attestation and understanding
All workforce members attest to the policies that apply to them at hire, every [12] months, and within [30] days of a material change. Role-based staff (for example, system administrators, developers and finance staff) attest to the standards for their role. The [GRC Manager] reports attestation rates to the [Executive Risk Committee] [quarterly], reviews policy portal usage and help desk questions to find documents that are unclear or unused, and works with [Security Awareness] to test understanding of high-risk requirements.
8. Review
Policies are reviewed at least every [12] months; standards and procedures at least every [12 / 24] months. A review is also required within [90] days of: a new or amended law, regulation or contract requirement; a significant incident; an audit, examination or test finding against the document; a merger, acquisition or divestiture; a major technology change; or [five] or more active exceptions against the same requirement. Each review is recorded with the reviewer, date and outcome, including "no change."
9. Compliance and enforcement
Compliance with governance documents is monitored by [Security GRC] and tested by [Internal Audit]. Violations are handled under [the Workforce Discipline Policy], applied consistently regardless of rank. Requests not to meet a requirement follow the Exception and Risk Acceptance Procedure.
10. Retention and retirement
Every approved version, with approval records and attestation records, is kept for [six] years after it is superseded or retired, or longer if a law, regulation, contract or legal hold requires. When a document is retired, the register records where its requirements moved.
11. Reporting
The [CISO] reports to the [Risk Committee] at least [annually] on the state of the governance documents: documents past their review date, attestation rates, open exceptions and significant changes made during the year.
Owner: [CISO]. Approved by: [Executive Risk Committee] on [date]. Version [1.0]. Next review: [date].
Sample: Exception and Risk Acceptance Procedure
Sample only. Adapt it to your organization's risks, systems and laws, have qualified people review it, and do not treat it as a substitute for a risk assessment.
Follows: NIST CSF 2.0, GV.RM (risk management strategy) and ID.RA-06 (risk responses chosen, tracked and communicated) (Feb. 26, 2024); NIST SP 800-37 Rev. 2 (risk acceptance by an authorizing official); 16 CFR 314.4(b)(1) (criteria for mitigating or accepting risks); 45 CFR 164.306(d)(3) (documenting addressable specifications); 23 NYCRR 500.17(b) (acknowledgment of noncompliance); PCI DSS v4.0.1 compensating controls (described); Idaho Technology Authority G120 (exemption process).
1. Purpose
This procedure sets how [Organization name] grants time-limited exceptions to its information security and privacy standards and records decisions to accept risk, so that every known gap is visible, rated, covered by other controls where possible, approved at the right level and closed or renewed on purpose.
2. What may not be excepted
No exception may be granted from a requirement of law, regulation or a binding contract. Where such a requirement is not met, the [Chief Compliance Officer] and [CISO] record a compliance gap with a remediation plan, decide with the [general counsel's office] whether any report, certification or customer notice must reflect it, and report it to the [Risk Committee]. No exception may be granted that would allow concealment of an incident or a false statement in any filing, certification or customer representation.
3. Request
The business or system owner submits a request in [the GRC platform] stating: the requirement not met; the systems, data, processes or people affected; the business reason; the period requested; compensating controls proposed; and the remediation plan with dates and funding.
4. Assessment
[Security GRC] rates the inherent and residual risk using the enterprise risk scale, confirms the compensating controls are in place and who will maintain them, checks whether the gap affects any regulatory certification or contract representation, and recommends approval, changes or denial within [10] business days.
5. Approval levels and terms
| Residual risk | Approver | Maximum term |
|---|---|---|
| Low | [GRC Manager] and system owner | [12] months |
| Medium | [CISO] and accountable business leader | [12] months |
| High | [Executive Risk Committee] after [CISO] review; reported to the [Risk Committee] of the Board | [6] months, renewable once |
| Critical | Not eligible; treated as an incident or compliance gap with an approved remediation plan | Not applicable |
The requester may not approve their own request. Approvers record the decision, the reason and any conditions.
6. Register and monitoring
All exceptions and risk acceptances are recorded in the exception register with owner, rating, compensating controls, approver, approval date, expiry date and remediation status. [Security GRC] reviews the register [monthly]; the [Executive Risk Committee] receives a [quarterly] report of open exceptions by rating, exceptions past expiry and exceptions renewed more than once; the [Risk Committee] receives the same report [at each meeting / semiannually].
7. Expiry and renewal
[Security GRC] notifies the owner [60] days before expiry. A renewal requires a new request with a progress report on remediation and is approved one level higher than the original. An expired exception that is neither renewed nor closed is reported as a policy violation.
8. Changes to production systems
Remediation that changes production systems follows the [Change Management Standard]: the change is planned, tested and has a documented way back.
Owner: [CISO]. Approved by: [Executive Risk Committee] on [date]. Version [1.0]. Next review: [date].
Pictures, Numbers and Infographics
The key numbers in this lesson, with their sources:
| Number | What it is | Source |
|---|---|---|
| 15 | Topic areas New York's 23 NYCRR 500.3 requires written policies to address | 23 NYCRR 500.31 |
| At least annually | How often New York requires a senior officer or the board to approve the policies | 23 NYCRR 500.31 |
| April 15 | Annual NYDFS certification or acknowledgment of noncompliance, signed by the top executive and the CISO | 23 NYCRR 500.17(b)9 |
| 5 years | Records supporting the NYDFS filing must be kept | 23 NYCRR 500.17(b)(3)9 |
| 6 years | HIPAA policies and required records must be kept from creation or last effective date | 45 CFR 164.316(b)16 |
| 500 consumers; 30 days | FTC notice threshold and deadline for a Safeguards Rule notification event | 16 CFR 314.4(j)2 |
| 6 months | Maximum interval between Safeguards Rule vulnerability assessments | 16 CFR 314.4(d)(2)2 |
| 24 months; 90 days | Maximum Idaho ITA exemption term; renewal request lead time | ITA Guideline G1205 |
| 14 days; 4 weeks | ITA comment period on draft documents; lead time before the committee meeting | ITA Policy P101523 |
| 7 | Minimum elements of an effective compliance and ethics program | USSG section 8B2.1(b)3 |
| 3 points | Culpability score reduction for an effective compliance program | USSG section 8C2.5(f)3 |
| 120 days | Window to self-report after an internal whistleblower report and keep a declination presumption (Criminal Division, Aug. 2024) | Justice Department31 |
| 34 | Control domains in the Secure Controls Framework, release 2026.2 (July 7, 2026) | SCF Council8 |
| About 27,000; 48; $170,221,335.88 | SEC whistleblower tips, individuals awarded, and amount paid in fiscal year 2025 | SEC FY2025 annual report33 |
| 10 to 30 percent; $1 million | SEC award range and sanctions threshold | SEC FY2025 annual report33 |
| 15 to 25 percent; 25 to 30 percent | False Claims Act relator share with and without government intervention | 31 U.S.C. 3730(d)34 |
| $9 million; $2.61 million | Aerojet Rocketdyne settlement and relator's share, July 8, 2022 | Justice Department35 |
| $1.25 million; $250,000 | Penn State settlement and relator's share, Oct. 22, 2024 | Justice Department13 |
| $875,000; $201,250 | Georgia Tech Research Corporation settlement and relators' share, Sept. 30, 2025 | Justice Department14 |
| $2,042,518; $375,823 | Honeywell Aerospace settlement and relator's share, Sept. 1, 2026 | Justice Department, as summarized by a law firm37 |
| 18; at least $84,271,106 | Civil Cyber-Fraud Initiative settlements and total through Sept. 1, 2026 (private count) | Settlement tracker36 |
| $3 million; $5 million; $1.5 million | NYDFS penalties against National Securities (2021), Carnival (2022) and Residential Mortgage Services (2021), each involving an improper compliance certification | NYDFS orders as summarized113839 |
| $80 million | OCC civil money penalty against Capital One for risk assessment failures before a cloud migration, Aug. 6, 2020 | OCC40 |
Common mistakes
Certifying compliance that is not true
New York's annual filing has produced a series of orders in which a company certified compliance with Part 500 and the Department of Financial Services later found it had not complied. National Securities Corporation paid $3 million in April 2021 after certifying compliance for 2018 despite failures in multifactor authentication and breach notice.11 Residential Mortgage Services paid $1.5 million in March 2021; its 2020 certification of compliance for 2019 was found inaccurate after an employee's email account was taken over through phishing, the company did not report the event to DFS within 72 hours, and it lacked a full risk assessment. DFS learned of the attack during a routine examination.39 Carnival Corporation paid $5 million in June 2022 and gave up its New York insurance producer license after four cybersecurity events from 2019 to 2021, including two ransomware attacks; DFS found that Carnival had not required multifactor authentication for email, had not trained staff adequately, reported the first event 10 months late, and had improperly certified compliance.38 The lesson for the April 15 filing: if the evidence does not support a certification, file the acknowledgment of noncompliance with a remediation timeline, which the rule expressly allows.9
Reporting scores and dates that do not match the systems
Georgia Tech's alleged score of 98 for a campus-wide environment that did not exist, Penn State's alleged inaccurate dates for closing control gaps, and MORSECORP's reported score of 104 against a later consultant score of -142 are all failures of review before submission.141336 A compliance step that ties every submitted score to a defined system boundary, a current system security plan and evidence would have caught each one.
No written policies, no senior owner
The FTC's 2022 action against the alcohol delivery company Drizly and its chief executive alleged that the company did not "develop adequate written security policies, or train employees on those procedures," and did not "put a senior executive in charge of ensuring that the company was keeping its data secure." The order applies to the CEO personally and follows him to future companies that collect information on more than 25,000 individuals.41 A policy library with named owners and a named senior executive is the most basic answer to both allegations.
Doing the risk assessment after the decision
The OCC's $80 million penalty against Capital One on Aug. 6, 2020 cited the bank's "failure to establish effective risk assessment processes prior to migrating significant information technology operations" to the public cloud, and its "failure to correct the deficiencies in a timely manner."40 Policy should require the risk assessment, the applicable standards and any exceptions to be settled before a major technology change goes live, not afterward.
Exceptions that never end
A recurring audit finding in large organizations is the exception approved years ago and renewed by habit, or the known gap never entered in the register. Both leave leadership believing a standard is met. Expiry dates, renewal at a higher level and quarterly reporting of overdue and repeatedly renewed exceptions are the controls; the sample procedure above includes them.
Describing a program that does not operate
The SEC's case against SolarWinds and its CISO, filed Oct. 30, 2023, centered on a public security statement that the company followed the NIST Cybersecurity Framework and had strong access controls, set against internal assessments. A federal judge dismissed most of the case on July 18, 2024 but let the claims about the security statement proceed, and the SEC dismissed the remaining claims with prejudice on Nov. 20, 2025.42 Whatever one thinks of the case's end, it showed that public statements drawn from policy documents will be compared with internal records. 108 lesson 4 covers the case in full.
Confidentiality terms that discourage reports
Severance agreements, nondisclosure agreements and incident response confidentiality instructions that appear to bar employees from talking to regulators create risk under the SEC's whistleblower rules and undercut the anti-retaliation element of the Sentencing Guidelines.333 Every such template should carry a plain carve-out for reports to government agencies.
Questions people ask
How long should a security policy be?
The board-approved master policy should be short enough for directors to read and discuss: a few pages that state commitments, roles, authority and the documents below it. Detail belongs in standards. Many large organizations have a master policy under 10 pages, 10 to 20 topic policies, and many more standards and procedures. A single 300-page policy approved by the board each year usually means nobody reads it, and any small change requires board approval.
Does the board have to approve every security policy?
No. Rules that require approval usually name the board or a senior officer for the program or top-level policy: New York allows "a senior officer or the covered entity's senior governing body," and the banking guidelines give the board or a committee responsibility for the program.115 The board should approve the master policy and the risk appetite, and delegate the rest by name in that policy. What matters is that the delegation is written and the approvals are recorded.
Is annual policy attestation required by law?
Few rules use the word, but many require the organization to communicate policies and train people, and auditors ask for proof. The Sentencing Guidelines require periodic, practical communication of standards and procedures; PCI DSS requires security awareness training; HIPAA requires a security awareness and training program; and the New York rule requires annual cybersecurity awareness training.3171627 Attestation is the most common way to show the communication happened.
Do we need a GRC platform, or will spreadsheets do?
Spreadsheets work for a smaller program with a disciplined owner, version control and restricted editing. As the number of regulations, audits and control owners grows, spreadsheets break down: evidence is duplicated, versions conflict and nobody can produce a history. Most large organizations move to a platform, but the process and control library should come first; a tool configured around an undefined process automates confusion.
Will auditors accept our unified control framework instead of their own checklist?
Usually, if the mapping is documented and the evidence meets their standard. A PCI assessor still tests against PCI requirements, a SOC 2 auditor against the Trust Services Criteria and a CMMC assessor against NIST SP 800-171, but each can rely on the same tested control when the control is written to the strictest requirement and the evidence fits their method. Share the mapping with the auditor at planning, not during fieldwork.
If we are compliant, are we secure?
No. Compliance shows that the organization meets a set of minimum requirements, as tested at a point in time or over a period. Attackers do not check certifications. Organizations with clean audits have suffered major breaches, and lesson 5 discusses breached companies with clean SOC 2 reports. A good compliance program supports security by making controls visible and tested; it does not replace risk-based judgment about what the organization needs.
What should happen when an employee says a past certification was wrong?
Treat it as a serious report: log it through the ethics or compliance office, protect the employee from retaliation, investigate with people independent of the program in question, involve the general counsel's office, and if the certification was wrong, correct it with the agency that received it. The Justice Department's policies give credit for prompt self-disclosure and remediation, and its 2024 policy gives a company 120 days after an internal whistleblower report to self-report and keep a declination presumption in Criminal Division matters.31 Delay makes every option worse.
Hard questions
Do personal certifications by executives make programs better, or only create scapegoats?
The case for: requiring the chief executive and the CISO to sign an annual compliance filing, as New York does, or a senior official to affirm CMMC compliance each year, forces senior attention, budgets and evidence. People read what they must sign. The Sarbanes-Oxley certifications by CEOs and CFOs, adopted in 2002, are often credited with improving financial controls for the same reason.
The case against: security is never perfect, the rules are long and partly subjective, and a personal signature turns a judgment call into potential personal liability. CISOs say this deters qualified people from taking the job and pushes them to hedge. The Uber and SolarWinds cases made many security leaders more cautious about what they put in writing.
What the evidence shows: the enforcement record so far targets certifications contradicted by facts the signers knew or should have found, such as missing multifactor authentication or unreported incidents, rather than good-faith judgments that proved wrong.113839 The New York rule's option to file an acknowledgment of noncompliance with a remediation plan gives signers an honest alternative.9 There is no published study measuring whether these certifications reduce breaches.
Should cyber compliance report to the CISO or be independent?
For keeping it with the CISO: the people who understand the controls are in the security organization; separating them adds handoffs and cost and can produce a compliance team that checks boxes without understanding systems.
For independence: a function that reports to the person whose program it checks has a conflict. The Justice Department asks about compliance's autonomy, data access and direct access to the board, and the three lines model separates management of risk from oversight of it.4
What the evidence shows: there is no rule requiring one structure, and large banks and insurers often split the work, keeping technical GRC with the CISO and regulatory oversight in enterprise compliance. The documented failures (false certifications, scores for systems that did not exist) are failures of review and challenge, which an independent check is designed to catch. Size and regulatory exposure should drive the choice.
Is the False Claims Act the right tool for cybersecurity?
For: taxpayers pay for security that contractors promise. The act lets insiders who know the truth bring cases the government would never find, and the settlements have been modest relative to contract values while sending a clear message that scores and certifications must be true. Defense officials have said the information at stake, controlled unclassified information about weapons and research, is exactly what adversaries seek.
Against: security requirements such as NIST SP 800-171 are long, technical and partly subjective; treble damages and per-claim penalties are severe for what may be honest disagreements about whether a control is "implemented"; and universities and small suppliers may leave federal work rather than carry the risk. Some argue that contract remedies and CMMC assessments, not fraud law, should handle shortfalls. The Department of War's July 2026 suspension of CMMC Phase II and its review of program costs reflect concern about the burden on the defense industrial base.30
What the evidence shows: the cases settled to date involve allegations of misstatement, such as inflated scores, dates not met, a nonexistent environment, rather than technical disagreement alone, and most were brought by insiders.131437 Whether the cases improved security across contractors has not been measured in any public study this lesson found.
Does compliance automation make programs stronger or turn them into checklists?
For automation: continuous monitoring finds failed controls in days, cuts audit preparation time and produces better evidence than screenshots collected once a year.
Against: automation measures what is easy to measure; vendors sell "audit-ready in weeks" packages that can produce a SOC 2 report for a program with little substance; and a green dashboard can stop people from asking whether the controls address the organization's real risks.
What the evidence shows: neither side has strong public data. Regulators continue to ask for evidence of operation and of root-cause correction, which automation can supply only for the controls it covers.4 The prudent course is to automate collection but keep human design review, sampling and internal audit.
For reporters and advocates
What public records exist.
- SEC filings on EDGAR. Every public company's 10-K Item 1C describes its cyber risk management processes and board oversight; proxy statements describe committee charters. Compare what a company says its policies require with what happened in its next Form 8-K Item 1.05 incident filing.21
- NYDFS enforcement. The Department of Financial Services posts consent orders, which often recite what the company certified and what DFS found.11
- Justice Department releases and court dockets. False Claims Act settlements are announced by the Civil Division and the U.S. Attorney's offices; unsealed qui tam complaints, such as U.S. ex rel. Decker v. Pennsylvania State University, No. 2:22-cv-03895 (E.D. Pa.), are on PACER and often free on CourtListener.13
- SEC whistleblower annual reports to Congress give tip counts and award totals by year.33
- FTC and OCC orders. FTC orders such as Drizly's and OCC enforcement actions such as Capital One's are public and often describe the policy and governance failures in detail.4140
- Idaho records. The ITA's policies, standards and guidelines and its exemption process are published online, and ITA and IT Leadership Council meeting minutes and exemption decisions can be requested under the Idaho Public Records Act, Idaho Code Title 74, Chapter 1, subject to its exemptions for security-sensitive records.225 The Legislative Services Office's audit division publishes audits of state agencies.
Questions to ask an organization or agency.
- Who approved your information security policy, and when? When was each standard last reviewed?
- How many policy exceptions are open, how many are high risk, and how many are past their expiry date? Who approved the oldest one?
- Who signs your annual certifications (NYDFS, CMMC, the Safeguards Rule board report), and what sub-certifications and evidence support them?
- Does compliance report to the security chief, or independently? Can it reach the board without management's permission?
- For a public agency: which statewide policies does the agency have an exemption from, for how long, and why?
- For a contractor: what SPRS score has the company reported, for which systems, and when was it last assessed by someone outside the company?
- How does an employee report a security concern without going through the security team, and what happened to the last few reports?
Cautions. A settlement is not an admission of liability, and a consent order's findings are the regulator's. A policy exception is not a violation; an undocumented gap may be. Compare dates carefully: a certification is judged by what was true in the period it covers.
Where to go next
- NIST CSF 2.0 Reference Tool and OLIR catalog: Free. Browse CSF 2.0 outcomes with informative references and published mappings to other standards; for compliance teams building or checking a control library.
- NIST SP 1347, CSF 2.0 Informative References Quick-Start Guide: Free. A short guide to what informative references are and how to download and use them, with sample use cases; start here before building mappings.
- DOJ Evaluation of Corporate Compliance Programs: Free. The questions prosecutors ask about any compliance program; compliance officers and boards can use it as a self-assessment checklist.
- U.S. Sentencing Guidelines, Chapter 8 (organizations): Free. Section 8B2.1 sets the seven elements of an effective compliance and ethics program; read with the commentary.
- Secure Controls Framework: Free download from a private council. A control library with mappings to many laws and frameworks; useful to start a unified control framework, but check mappings for your highest-risk rules.
- NYDFS Cybersecurity Resource Center: Free. Part 500 text, FAQs, filing instructions for the April 15 certification, and enforcement actions; for New York licensees and anyone studying certification practice.
- Idaho Technology Authority policies, standards and guidelines: Free. Idaho's statewide IT policy library with its development process (P1015) and exemption process (G120); a public model of a policy hierarchy for agencies, auditors and reporters.
Still being checked (9 items)
These points could not be fully confirmed against a primary source when this lesson was written and fact checked in November 2026. The text above states each one as uncertain.
- CMMC Phase II status after Aug. 20, 2026: the Reform Task Force's recommendations and any restart date could not be confirmed; the suspension is described from law firm summaries of the July 13, 2026 memoranda, not the memoranda themselves.
- The Health Net and Centene, Raytheon, RTX and Nightwing, and Illumina settlement details, and the count of 18 settlements totaling at least $84,271,106, come from a private settlement tracker; the Justice Department releases for those cases were not re-read.
- The Honeywell Aerospace settlement (Sept. 1, 2026) is stated from a law firm summary; the Justice Department release was blocked to automated reading on Nov. 16, 2026. The amount ($2,042,518), conduct period (April 2020 to December 2023), relator (Rachel Tenney) and relator share ($375,823) were confirmed against a second, independent news report (KTAR, Sept. 5, 2026); the 2022 filing year and Aug. 18, 2026 partial intervention were not independently confirmed.
- The National Securities, Residential Mortgage Services and Carnival NYDFS orders are stated from law firm summaries, not the consent orders themselves.
- PCI DSS v4.0.1 requirement numbers and wording (12-month policy review, Requirement 8.4, customized approach) are described from the published standard and were not re-read for this lesson.
- 23 NYCRR 500.12, 500.13 and 500.14 are described as verified for module 108; only 500.3, 500.5, 500.9 and 500.17 were re-read for this lesson.
- The NYDFS Cybersecurity Resource Center URL in Learn more was not opened for this lesson.
- The SEC FY2025 whistleblower report date is given as the fiscal year end; its release date was not confirmed. Its figures (about 27,000 tips; more than $60 million to 48 individuals; $170,221,335.88 paid from the fund) were confirmed against the report PDF on Nov. 16, 2026.
- IRS Publication 1075, the FBI CJIS Security Policy and CMS rules are mentioned generally without current version numbers.
Last verified November 15, 2026. General education, not legal, compliance, audit or other professional advice; rule versions and dates are stated in the text, and the sources used are listed in the Notes and Sources below. Check the regulator's current text before relying on any point.
Notes
- New York State Department of Financial Services, 23 NYCRR 500.3, Cybersecurity policy (written policies "approved at least annually by a senior officer or the covered entity's senior governing body," topics (a) to (o)), Part 500 as amended effective Nov. 1, 2023, Cornell Legal Information Institute copy (read Nov. 15, 2026). Read it online. ↩
- Federal Trade Commission, Standards for Safeguarding Customer Information, 16 CFR 314.4, as amended at 88 FR 77508 (Nov. 13, 2023): (b)(1) written risk assessment, (c)(5) multifactor authentication, (d)(2) annual penetration testing and vulnerability assessments at least every six months, (i) written report to the board at least annually, (j) notice to the FTC within 30 days for events involving at least 500 consumers, eCFR (read Nov. 15, 2026). Read it online. ↩
- U.S. Sentencing Commission, Guidelines Manual (2026), Chapter 8, section 8B2.1 (Effective Compliance and Ethics Program), including (a), (b)(1) to (7) and (c), and section 8C2.5(f) (three-point culpability reduction and its exclusions), annotated online edition; the historical notes list no amendment to 8B2.1 in 2025 or 2026 (read Nov. 15, 2026). Read it online. ↩
- U.S. Department of Justice, Criminal Division, Evaluation of Corporate Compliance Programs (updated September 2024): the three fundamental questions; policies and procedures; risk assessment; autonomy and resources; emerging technology; anti-retaliation; continuous improvement (read Nov. 15, 2026). Read it online. ↩
- Idaho Technology Authority, Guideline G120, Exemption Process (effective Feb. 19, 2003; footer revision June 30, 2004; revision history entry April 20, 2021; web page modified July 30, 2026): exemptions "not to exceed twenty-four (24) months"; renewal request at least 90 days before expiry; policy exemptions to the ITA, standards exemptions to the IT Leadership Council (read Nov. 15, 2026). Read it online. ↩
- National Institute of Standards and Technology, NIST IR 8477, Mapping Relationships Between Documentary Standards, Regulations, Frameworks, and Guidelines: Developing Cybersecurity and Privacy Concept Mappings, final Feb. 26, 2024 (draft Aug. 17, 2023); describes concept crosswalks, set theory relationship mappings and supportive relationship mappings (read Nov. 15, 2026). Read it online. ↩
- NIST Special Publication 1347, NIST Cybersecurity Framework (CSF) 2.0: Informative References Quick-Start Guide, final, announced Aug. 2026, on obtaining references by direct download, the CSF 2.0 Reference Tool and the OLIR program (read Nov. 15, 2026). Read it online. ↩
- Secure Controls Framework Council, "SCF 2026.2" newsletter (release dated July 7, 2026): 34 control domains including a new Quantum Security domain; set theory relationship mappings (STRM) published for mapped sources. A free download from a private organization (read Nov. 15, 2026). Read it online. ↩
- 23 NYCRR 500.17(b), annual certification of material compliance or acknowledgment of noncompliance due April 15, signed by the highest-ranking executive and the CISO, supporting records kept five years, Cornell Legal Information Institute copy (read Nov. 15, 2026). Read it online. ↩
- 32 CFR 170.22, Affirmation (CMMC Program rule, effective Dec. 16, 2024): affirmations by the Affirming Official "after every assessment, including POA&M closeout, and annually thereafter," entered in SPRS, eCFR (read Nov. 15, 2026). Read it online. ↩
- Akin Gump, "New York Department of Financial Services Issues Millions of Dollars in Penalties Signaling Increased Cybersecurity Enforcement" (2021), summarizing the April 14, 2021 NYDFS consent order with National Securities Corporation ($3 million; 2018 certification of compliance despite multifactor authentication and notice failures) and the 2021 charges against First American Title. Secondary summary of NYDFS orders (read Nov. 15, 2026). Read it online. ↩
- U.S. Department of Justice, "Deputy Attorney General Lisa O. Monaco Announces New Civil Cyber-Fraud Initiative," Oct. 6, 2021 (read Nov. 15, 2026). Read it online. ↩
- U.S. Attorney's Office, Eastern District of Pennsylvania, "Penn State Agrees to Pay $1.25 Million to Resolve False Claims Act Allegations Relating to Non-Compliance with Contractual Cybersecurity Requirements," Oct. 22, 2024; U.S. ex rel. Decker v. Pennsylvania State University, No. 2:22-cv-03895 (E.D. Pa.); relator's share $250,000 (read Nov. 15, 2026). Read it online. ↩
- U.S. Department of Justice, "Georgia Tech Research Corporation Agrees to Pay $875,000 to Resolve Civil Cyber-Fraud Litigation," Sept. 30, 2025 (Astrolavos Lab; score of 98 submitted December 2020; relators Christopher Craig and Kyle Koza, share $201,250; government intervened August 2024) (read Nov. 15, 2026). Read it online. ↩
- Interagency Guidelines Establishing Information Security Standards, 12 CFR Part 30, Appendix B (OCC version; parallel guidelines apply to other federally supervised banks), sections III.A (board involvement) and III.F (report to the board at least annually), eCFR (read Nov. 15, 2026). Read it online. ↩
- 45 CFR Part 164, Subpart C (HIPAA Security Rule), sections 164.306(d)(3) (addressable implementation specifications), 164.308(a)(1) (risk analysis), 164.308(a)(5) (training), 164.308(a)(8) (evaluation) and 164.316(b) (documentation kept six years, available, reviewed periodically), eCFR (read Nov. 15, 2026). Read it online. ↩
- PCI Security Standards Council, "Just Published: PCI DSS v4.0.1," June 11, 2024, and the PCI DSS v4.0.1 standard (Requirement 12 policy and program requirements, including review of the information security policy at least once every 12 months; Requirement 8.4 multifactor authentication; compensating controls and the customized approach). The standard is licensed and is described, not reproduced; requirement wording was not re-read for this lesson (read Nov. 15, 2026). Read it online. ↩
- NIST SP 800-53 Rev. 5 (Update 1), Security and Privacy Controls for Information Systems and Organizations: each control family begins with a "Policy and Procedures" control (for example AC-1, IR-1) with organization-defined review frequencies (read Nov. 15, 2026). Read it online. ↩
- NIST CSWP 29, The NIST Cybersecurity Framework (CSF) 2.0, Feb. 26, 2024: Govern function, Policy category (GV.PO-01, GV.PO-02), Risk Management Strategy (GV.RM), ID.RA-06, PR.AA-03 and informative references (read Nov. 15, 2026). Read it online. ↩
- Idaho Technology Authority, Enterprise Policy P4140, Cybersecurity Framework (authority Idaho Code 67-831 through 67-833; established Oct. 17, 2001; revised Aug. 20, 2024): agencies "shall adopt the current NIST Cybersecurity Framework as the basis for assessing their cybersecurity programs" (quoted as verified for module 108, Nov. 13, 2026) (read Nov. 15, 2026). Read it online. ↩
- U.S. Securities and Exchange Commission, Release No. 33-11216, "Cybersecurity Risk Management, Strategy, Governance, and Incident Disclosure" (July 26, 2023; 88 FR 51896, Aug. 4, 2023), adding Regulation S-K Item 106 (17 CFR 229.106), reported in Form 10-K Item 1C, and Form 8-K Item 1.05 (read Nov. 15, 2026). Read it online. ↩
- Idaho Technology Authority, "Policies, Standards and Guidelines" index (P policy, S standard, G guideline; 1000 general series; 4000 and 4500 security policies, including P4110, P4130, P4140, P4502, P4520 and P4590) (read Nov. 15, 2026). Read it online. ↩
- Idaho Technology Authority, Policy P1015, IT Policies, Standards, and Guidelines Development (established Aug. 5, 2024; page updated Oct. 8, 2026): drafting on ITA templates, accessibility check, four-week lead time, 14-day comment period, subcommittee approval and ITA review (read Nov. 15, 2026). Read it online. ↩
- HHS Office for Civil Rights, "HIPAA Security Rule Notice of Proposed Rulemaking to Strengthen Cybersecurity for Electronic Protected Health Information," fact sheet (published Jan. 6, 2025, 90 FR 898, RIN 0945-AA22). Still a proposal; the Unified Agenda listed it among long-term actions as checked for module 108 on Nov. 14, 2026 (read Nov. 15, 2026). Read it online. ↩
- NIST, National Online Informative References (OLIR) Program and the CSF 2.0 Reference Tool, Computer Security Resource Center (read Nov. 15, 2026). Read it online. ↩
- U.S. Securities and Exchange Commission, Form 10-K, General Instructions A(2) (due 60 days after fiscal year end for large accelerated filers, 75 days for accelerated filers, 90 days for other registrants) (read Nov. 15, 2026). Read it online. ↩
- 23 NYCRR Part 500 as amended effective Nov. 1, 2023: 500.4 (CISO report to the board at least annually), 500.5 (penetration testing "at least annually"), 500.9(a) (risk assessment "reviewed and updated as reasonably necessary, but at a minimum annually"), 500.12 (multifactor authentication for any individual accessing information systems, effective Nov. 1, 2025), 500.13 (asset inventory) and 500.14 (annual training). Sections 500.5 and 500.9 re-read for this lesson; 500.12, 500.13 and 500.14 are described as verified for module 108 (read Nov. 15, 2026). Read it online. ↩
- California Privacy Protection Agency, announcement of Office of Administrative Law approval of regulations on cybersecurity audits, risk assessments and automated decisionmaking technology (Sept. 23, 2025; effective Jan. 1, 2026), with compliance dates as summarized by Skadden, "California Finalizes CPPA Regulations" (Oct. 2025) (read Nov. 15, 2026). Read it online. ↩
- U.S. Securities and Exchange Commission, press release 2024-58, "SEC Adopts Rule Amendments to Regulation S-P to Enhance Protection of Customer Information," May 16, 2024 (89 FR 47688, June 3, 2024; compliance Dec. 3, 2025 for larger entities and June 3, 2026 for smaller entities), as verified for module 108 (read Nov. 15, 2026). Read it online. ↩
- Holland & Knight, "DOW Suspends CMMC Phase II Requirements" (July 31, 2026), describing the two Department of War memoranda of July 13, 2026 (CIO memorandum "Removing Barriers to Defense Industrial Base Expansion" creating the CMMC Reform Task Force and a 60-day review; implementation memorandum from the Under Secretary for Acquisition and Sustainment), the request for information closing Aug. 14, 2026, and the obligations that remain. Secondary summary; the memoranda were not read directly (read Nov. 15, 2026). Read it online. ↩
- U.S. Department of Justice, Criminal Division, "Corporate Enforcement" page: Evaluation of Corporate Compliance Programs (September 2024); department-wide Corporate Enforcement Policy (March 10, 2026); White-Collar Enforcement Plan, revised self-disclosure policy and monitor selection memorandum (May 2025); Crediting Policy (June 2025); Corporate Whistleblower Awards Pilot Program (August 2024, updated May 2025); temporary amendment on internal whistleblower reports and 120-day self-reporting (August 2024) (read Nov. 15, 2026). Read it online. ↩
- Mayer Brown, "DOJ Releases First-Ever Department-Wide Corporate Enforcement Policy for All Criminal Matters," March 12, 2026 (read Nov. 15, 2026). Read it online. ↩
- U.S. Securities and Exchange Commission, Annual Report to Congress on the Whistleblower Program, Fiscal Year 2025: about 27,000 tips; more than $60 million awarded to 48 individuals; $170,221,335.88 paid from the Investor Protection Fund; awards of 10 to 30 percent where sanctions exceed $1 million (read Nov. 15, 2026). Read it online. ↩
- 31 U.S.C. 3730, Civil actions for false claims: (d) relator share of 15 to 25 percent with government intervention and 25 to 30 percent without; (h) relief from retaliation, including reinstatement and "2 times the amount of back pay," within three years. Liability and penalties are in 31 U.S.C. 3729. Cornell Legal Information Institute copy (read Nov. 15, 2026). Read it online. ↩
- U.S. Attorney's Office, Eastern District of California, "Aerojet Rocketdyne Agrees to Pay $9 Million to Resolve False Claims Act Allegations of Cybersecurity Violations in Federal Government Contracts," July 8, 2022 (relator Brian Markus, share $2.61 million; settled on the second day of trial) (read Nov. 15, 2026). Read it online. ↩
- The Defense Compliance Report, "Cyber False Claims Act Settlements: DOJ Settlement Tracker (2021 to 2026)," listing 18 Civil Cyber-Fraud Initiative settlements through Sept. 1, 2026 totaling at least $84,271,106, including Health Net Federal Services and Centene ($11,253,400, Feb. 18, 2025), MORSECORP ($4.6 million, March 26, 2025), Raytheon, RTX and Nightwing ($8.4 million, May 1, 2025) and Illumina ($9.8 million, July 31, 2025). A private tracker; the underlying Justice Department releases for these four were not re-read for this lesson. MORSECORP's admitted facts (SPRS score of 104; consultant score of -142; non-FedRAMP email host) were verified for module 108 on Nov. 13, 2026 (read Nov. 15, 2026). Read it online. ↩
- Nixon Peabody, "Honeywell settlement shows FCA risk for NIST 800-171 gaps" (Sept. 9, 2026), summarizing the Justice Department's Sept. 1, 2026 announcement: $2,042,518; conduct April 2020 to December 2023; relator Rachel Tenney (qui tam filed 2022; partial intervention Aug. 18, 2026; share $375,823). The department's own release could not be retrieved for this lesson (read Nov. 15, 2026). Read it online. ↩
- Hunton Andrews Kurth, "NYDFS Imposes Fine of $5 Million on Carnival for Cybersecurity Breaches" (July 1, 2022), on the NYDFS settlement announced June 24, 2022 (four events 2019 to 2021; first reported 10 months late; no MFA in email; inadequate training; improper certification; insurance producer license surrendered). Secondary summary (read Nov. 15, 2026). Read it online. ↩
- Crowell and Moring, "Recent Cyber Incidents Likely to Expose Lacking Substance in Risk Assessments and Self-Certifications" (2021), summarizing the March 2021 NYDFS consent order with Residential Mortgage Services ($1.5 million; inaccurate 2020 certification for 2019). Secondary summary (read Nov. 15, 2026). Read it online. ↩
- Office of the Comptroller of the Currency, News Release 2020-101, "OCC Assesses $80 Million Civil Money Penalty Against Capital One," Aug. 6, 2020 (read Nov. 15, 2026). Read it online. ↩
- Federal Trade Commission, "FTC Takes Action Against Drizly and its CEO James Cory Rellas for Security Failures that Exposed Data of 2.5 Million Consumers," Oct. 24, 2022 (final order Jan. 10, 2023) (read Nov. 15, 2026). Read it online. ↩
- U.S. Securities and Exchange Commission, Litigation Release No. 26423, SEC v. SolarWinds Corp. and Timothy G. Brown, Nov. 20, 2025 (dismissal with prejudice; complaint filed Oct. 30, 2023, S.D.N.Y.; most claims dismissed July 18, 2024), as verified for module 108 (read Nov. 15, 2026). Read it online. ↩
Check yourself
1. Under New York's 23 NYCRR 500.3, how often must a covered entity's cybersecurity policies be approved, and by whom?
23 NYCRR 500.3 requires written policies 'approved at least annually by a senior officer or the covered entity's senior governing body.'
2. Which layer of governance documents states specific, measurable, mandatory rules such as 'MFA on all remote and administrative access'?
Standards carry out a policy with specific, testable requirements; guidelines are recommendations and procedures give the steps (NIST CSF 2.0 GV.PO and common practice described in this lesson).
3. What is the longest exemption the Idaho Technology Authority's Guideline G120 allows from a statewide IT policy or standard?
ITA Guideline G120 allows an exemption for a period 'not to exceed twenty-four (24) months,' with renewal requests due at least 90 days before expiry.
4. Who must sign New York's annual Part 500 certification or acknowledgment due April 15?
23 NYCRR 500.17(b)(2) requires signature by the highest-ranking executive and the CISO (or the senior officer responsible for the program if there is no CISO).
5. According to NIST IR 8477's set theory method, which relationship best describes most mappings between a regulation and a framework?
NIST IR 8477 (Feb. 26, 2024) defines set theory relationships; most regulation-to-framework pairs overlap without either fully covering the other, which is 'intersects with.'
6. How often must a Qualified Individual report in writing to the board under the FTC Safeguards Rule?
16 CFR 314.4(i) requires a written report 'regularly and at least annually' to the board or equivalent governing body.
7. Under the U.S. Sentencing Guidelines, what is the culpability score reduction for having an effective compliance and ethics program, absent the exclusions?
USSG section 8C2.5(f)(1) subtracts 3 points; it does not apply if the organization unreasonably delayed reporting or, generally, if high-level personnel were involved.
8. Under CMMC's 32 CFR 170.22, when must an Affirming Official affirm compliance in SPRS?
32 CFR 170.22(a) requires affirmations 'after every assessment, including POA&M closeout, and annually thereafter,' entered in SPRS. The July 13, 2026 Phase II suspension did not remove annual affirmations.
9. In the October 2024 Penn State False Claims Act settlement, who brought the case?
The Justice Department's Oct. 22, 2024 release names Matthew Decker, former CIO of the Applied Research Laboratory, as relator; he received $250,000 of the $1.25 million settlement.
10. What did the Justice Department's August 2024 temporary amendment allow a company that receives an internal whistleblower report?
The Criminal Division's August 2024 temporary amendment to its self-disclosure policy lets a company self-report within 120 days of an internal whistleblower report and still qualify for a declination presumption.
Terms in this lesson
- Policy
- A governance document that states what the organization requires and why, assigns responsibility and sets the authority of the documents below it; approved at a senior level.
- Standard
- A mandatory, specific and measurable rule that carries out a policy, such as a requirement for multifactor authentication on all remote access.
- Procedure
- Step-by-step instructions for the people who perform a control, owned by the team that does the work.
- Guideline
- Recommended practice that explains how to meet a standard well; not mandatory.
- Attestation
- A person's recorded confirmation that they have read and will follow the policies that apply to them, usually collected at hire, yearly and after material changes.
- Exception
- Formal, time-limited permission not to meet a specific requirement for a specific system, process or group, granted through a documented process and recorded in a register.
- Risk acceptance
- A decision by someone with the authority to do so that a remaining risk is within the organization's appetite and will be carried rather than reduced further.
- Compensating control
- A different control that reduces the same risk when a required control cannot be used; PCI DSS has a formal method for documenting one.
- Unified control framework
- A single library of control statements, each with an owner, test and evidence, mapped to every law, rule and standard it satisfies.
- Set theory relationship mapping
- The NIST IR 8477 method of describing how two requirements relate: subset of, superset of, equal, intersects with, or not related.
- GRC platform
- Governance, risk and compliance software that holds policies, the control library, risk register, evidence, findings and vendor records in one system.
- Continuous control monitoring
- Automated, frequent testing of controls, such as daily checks of MFA coverage or overdue vulnerabilities, whose results also serve as evidence.
- Sub-certification
- A written statement by a control owner or business leader that their controls operated as described, collected before executives sign a regulatory certification.
- Qui tam
- A False Claims Act suit filed by a private person (the relator) on the government's behalf; the relator receives a share of any recovery.
- Civil Cyber-Fraud Initiative
- A Justice Department effort announced Oct. 6, 2021 to use the False Claims Act against contractors and grantees that knowingly misrepresent cybersecurity or fail to meet cyber requirements.
Sources
- 23 NYCRR Part 500 (NYDFS Cybersecurity Regulation), sections 500.3 and 500.17 (2023-11-01) checked 2026-11
- FTC Safeguards Rule, 16 CFR 314.4 (2023-11-13) checked 2026-11
- U.S. Sentencing Guidelines Manual (2026), Chapter 8, annotated (2026-11-01) checked 2026-11
- DOJ Criminal Division, Evaluation of Corporate Compliance Programs (2024-09-23) checked 2026-11
- DOJ Criminal Division, Corporate Enforcement policies and pilot programs (2026-03-10) checked 2026-11
- NIST IR 8477, Mapping Relationships Between Documentary Standards, Regulations, Frameworks, and Guidelines (2024-02-26) checked 2026-11
- NIST SP 1347, CSF 2.0 Informative References Quick-Start Guide (2026-08-25) checked 2026-11
- NIST Cybersecurity Framework 2.0 (CSWP 29) (2024-02-26) checked 2026-11
- 32 CFR 170.22, CMMC affirmations (2024-12-16) checked 2026-11
- DOJ, Deputy Attorney General announces Civil Cyber-Fraud Initiative (2021-10-06) checked 2026-11
- DOJ, Penn State False Claims Act cybersecurity settlement (2024-10-22) checked 2026-11
- DOJ, Georgia Tech Research Corporation civil cyber-fraud settlement (2025-09-30) checked 2026-11
- SEC Whistleblower Program, FY2025 Annual Report to Congress (2025-09-30) checked 2026-11
- Idaho Technology Authority, P1015 (document development) and G120 (exemption process) (2024-08-05) checked 2026-11