AI agent assurance frameworks
Assurance frameworks differ on the only dimension buyers care about: whether an independent party can certify you against them. Most cannot. Knowing which is which prevents both wasted effort and claims that will not survive scrutiny.
Get an Agent Trust Gap BriefCertifiable, voluntary, or binding — the only distinction that matters
| Framework | Type | Certification | Bearing on agents |
|---|---|---|---|
| ISO/IEC 42001 | AI management system | Yes — accredited third-party audit | Management-system level. Proves you run a governed process, not that a given agent is safe. |
| NIST AI RMF | Voluntary framework | No certification exists | Govern / Map / Measure / Manage. Carries legal weight via Texas TRAIGA's affirmative defence. |
| SOC 2 | Attestation report | Yes — but not AI-specific | Security and availability controls. Buyers often accept it as a proxy; it does not evidence agent governance. |
| OWASP Agentic Top 10 (ASI01–ASI10) | Risk taxonomy | No | The nearest thing to shared vocabulary for agent risk. A checklist, not a certification. |
| EU AI Act | Binding regulation | Conformity assessment for high-risk | Legal obligation, not a framework you opt into. |
| ISO/IEC 27001 | Information security management | Yes | Foundational and agent-agnostic. Necessary, nowhere near sufficient. |
Verified 2026-08-07. Not legal advice. Confirm scope and current status with the issuing body before relying on any of this.
The three claims that do not survive scrutiny
- "NIST AI RMF certified." There is no certification body and no certificate. What exists is documented conformance, which is worth having and worth describing accurately.
- "SOC 2 covers our AI governance." SOC 2 attests to controls in defined trust criteria. Unless agent governance controls were explicitly in scope, it evidences something else.
- "ISO 42001 certified, therefore our agents are safe." ISO 42001 certifies a management system. It is meaningful and it is not a per-agent safety claim.
These matter because a buyer's security team will check, and an overstated certification claim damages trust more than having no certification at all.
What to do instead
Map controls to the frameworks that apply, keep the evidence, and describe your position precisely. "We map to NIST AI RMF and maintain documented conformance evidence" is both accurate and stronger than a claim that collapses on inspection. See the control crosswalk for how these map to each other.
Common questions
Which AI framework can you actually be certified against?
ISO/IEC 42001 offers accredited third-party certification of an AI management system. Most other AI frameworks, including NIST AI RMF and OWASP's agentic taxonomy, have no certification path.
Does NIST AI RMF certification exist?
No. There is no certification body and no certificate. Anyone selling NIST AI RMF certification is selling something that does not exist.
Is SOC 2 enough for AI governance?
Not on its own. It attests to controls within defined trust criteria. Unless agent governance controls were explicitly scoped in, it evidences a different thing.
Does ISO 42001 mean our agents are safe?
It means you operate a certified management system for AI. That is meaningful and it is not a per-agent safety claim — the distinction matters to any competent reviewer.
Which framework should we start with?
NIST AI RMF for structure, because it costs nothing and other regimes point at it. Add ISO 42001 if you need a certificate a buyer can verify.