This hub collects every article, methodology document, and analysis this publication has produced on AI agent certification. The framework evaluates autonomous agents across seven weighted dimensions. The supporting analysis addresses the questions that arise around it: how it relates to standards such as ISO 42001 and NIST AI RMF, where it differs from EU AI Act conformity assessment, how dynamic generative behaviour requires a different audit approach, and how certification results feed insurance underwriting.
The Agent Certified framework evaluates autonomous AI agents across Trust, Context, Distribution, Product, Governance, AI Integration, and Autonomy. These articles explain the methodology and its most recent revision.
How Agent Certified evaluates autonomous AI agents across Trust, Context, Distribution, Product, Governance, Integration, and Autonomy. Weights, signals, and what each dimension asks of an operator.
The published reference document for the first version of the Agent Certified framework. Seven dimensions, five tiers, scoring rubric, and assessment process. The document the first wave of assessments was conducted under.
The current published reference for the Agent Certified framework. Updated dimension weights, revised tier thresholds, expanded standards crosswalk, and the assessment process that law firms, insurers, and regulators cite.
Traditional software audits assume deterministic outputs. Generative AI agents do not produce them. This article addresses why dynamic behaviour requires a different audit methodology and what that methodology looks like.
A 2026 analysis of why generative AI agents need a different certification approach. Non-determinism, drift, and the audit methods that actually work for agents that do not produce the same output twice.
The Agent Certified framework sits alongside, not inside, the EU AI Act's conformity assessment procedure. These articles map the relationship between certification, compliance, and the main international standards.
Conformity assessment, voluntary certification, third-party assurance. Why the three are not the same thing under the EU AI Act and why the distinction matters for operators seeking both regulatory cover and insurer confidence.
A practical comparison of NIST AI RMF, ISO 42001, and the EU AI Act for European operators of AI agents. Where they overlap, where they differ, and what each one demands in terms of documentation and evidence.
A practical guide to implementing ISO/IEC 42001:2023 as an AI management system, with clause-by-clause mapping to EU AI Act obligations and Agent Certified dimensions.
AIUC-1 is the first AI insurance certification standard, but it was built for US legal context. European operators face EU AI Act obligations it does not address. This article maps what is missing.
Why independent AI agent certification under ISO 42001, AIUC-1, and Agent Certified translates into better insurance outcomes: lower premiums, faster placement, and broader cover.