AI Security and Resilience: Testing an AI system against adversarial input

Auditors ask about AI security and resilience early, because the answer tells them how the rest of the program is likely to look. The outcome is NIST AI RMF MEASURE 2.7.

What the framework says

AI system security and resilience – as identified in the MAP function – are evaluated and documented.

That is the text of NIST AI RMF MEASURE 2.7. Read it closely: it describes an outcome to achieve, not a product to buy, which is why two practices of the same size can satisfy it in different ways and both be right.

What this looks like in practice

In an assessment, NIST AI RMF MEASURE 2.7 is not one question. It resolves into several, and each one is really asking for different evidence. These are among the questions asked about it:

• Are this AI system's security and resilience evaluated and documented, against the adversarial threats relevant to its type, such as evasion, data poisoning or model extraction?
• Where this generative AI system's content is shared outside the organization, are content authentication methods, such as watermarking, provenance metadata or cryptographic signatures, applied and their reliability measured?
• Has this generative AI system been red-teamed for the generative AI attacks that reach the parts of it the organization controls, including direct and indirect prompt injection, poisoning of its retrieval or fine-tuning data and, where the organization hosts or fine-tunes the model, model extraction and membership inference, and for abuse to attack other systems, such as generating malicious code or phishing content?

Ownership usually sits with the security team, though the evidence is often produced by someone else, which is where the trail breaks. A control that works but has no owner tends to stop working the month the person who quietly maintained it changes roles.

Who this applies to

NIST AI RMF is voluntary rather than binding, and that changes how it should be used. Adopting it does not create a legal obligation, and declining it is not a violation. It earns its place when a customer contract names it, when it gives structure to AI governance work you already owe under a law or a contract, or when a recognized practice is worth having on record. Treat it as a way of organizing that work, not as a second regime.

Building it, in order

The sequence below is the order that produces evidence as a by-product rather than as a separate documentation exercise:

1. Identify the AI-specific threats relevant to this system.
2. Test security controls and resilience against those threats.
3. Record findings and remediation.
4. Identify outputs that are distributed externally or could be mistaken for human-made content.
5. Apply suitable provenance methods, watermarks, provenance metadata or signatures.
6. Measure how reliably they persist and verify, including false-positive and false-negative rates.

Where else this maps

The questions behind NIST AI RMF MEASURE 2.7 also map to NIST AI 600-1 2.8, 2.9, which is voluntary guidance rather than law; following it does not by itself establish compliance with any law. That matters for scoping: the same evidence can be written once and referenced from each, rather than rewritten per framework.

Evidence an assessor asks for

For this outcome the evidence types that satisfy it are configuration and report. The distinction matters more than it looks: a policy states what you intend to do, a procedure states how, and a record proves it happened on a date. Auditors ask for all three, and a practice that has written the first two often has nothing for the third.

Date every document and keep the superseded versions. The framework sets no retention period, so this is our recommendation rather than a requirement: keep each version for as long as the system it describes is in use and for a period after it is retired, because a question about a past decision is answered from the version that was in force when it was made.

Where this usually goes wrong

The usual gap is a control that exists in practice and nowhere in writing. Someone does the work, reliably, and it has never been written down, so the organization cannot demonstrate it and cannot notice when it stops. The fix is not more control, it is a dated record and a named owner.

What to do about it

• Produce the evidence, not the intention. For NIST AI RMF MEASURE 2.7 that means a configuration export or screenshot showing the setting live, dated and a dated report with the period it covers and who reviewed it. If the only evidence is that someone knows how this works, you have a practice and not a record.
• Put a name on it. The security lead should be recorded as the owner. Where the work is done by someone else, say so in the document: the split between who owns it and who performs it is where the trail usually breaks. A control with no owner stops working the month the person who quietly maintained it changes roles, and nobody notices.
• Set a review cadence and keep the evidence of each review. An annual date with a dated output is a common floor; what fails an assessment is a control implemented once and never re-checked against a changed environment.
• Do not treat this as a legal obligation. It is a voluntary framework. Adopt it where a customer contract or a regulator's guidance names it, or where it gives structure to AI governance work you owe for other reasons, and say in your own documentation which of those reasons applies.

The recurring lesson with AI security and resilience is that documentation is not the paperwork after the control. It is the part an assessor can read, and therefore the part that has to exist.

Related outcomes

• AI System Performance: Measuring performance on your own population
• AI Fairness and Bias: Testing for bias in a clinical population
• AI Metrics Selection: Choosing metrics that reveal harm rather than only accuracy

References

• MEASURE 2.7, in NIST AI Risk Management Framework 1.0 (NIST AI 100-1). This is a voluntary framework, not a legal requirement. The quoted statement is NIST's own, from the Core tables of NIST AI 100-1.
• Crosswalk references in this post are the mappings in our question bank, and are not verified against the published text of those standards.
• This post describes what a framework outcome asks for and what evidence documents it. It is not legal advice, and your counsel owns the conclusion. State law may impose additional obligations of its own.