This page explains who can be assessed under the Agent Certified methodology, what the assessment process involves, how institutional participants engage with this platform, and what we anticipate for a paid institutional partner tier in 2027. It should be read alongside our Editorial Standards.
Agent Certified is an independent methodology and assessment framework for autonomous AI agents operating in European organisations. We publish a seven-dimension methodology aligned with Regulation (EU) 2024/1689, ISO/IEC 42001:2023, and the NIST AI Risk Management Framework. Organisations that request an assessment and meet the criteria receive a published certification level. Carriers, investors, and procurement teams use our methodology as a reference for due diligence and underwriting decisions.
The methodology is published in full at methodology-v2.html. Any reader can inspect the evidentiary basis for any assessment outcome. This transparency is the foundation of the platform's value as a reference document for third party due diligence.
In 2026, all editorial coverage on this site, including articles, analysis, and framework comparisons, is produced on an editorial-only basis. No carrier, vendor, consultancy, or third party pays for placement, ranking, or favourable framing in our editorial content. Assessment services are provided at published rates and are independent of editorial coverage decisions.
This commitment is set out in detail in our Editorial Standards.
The Agent Certified assessment is available to any organisation that meets the following criteria.
An operator qualifies for assessment if it has deployed at least one autonomous AI agent in a production environment, meaning the agent is performing tasks with real consequences for real users or on real data, not in a test or sandbox environment only. Operators in pre-deployment stages may request a preliminary readiness review, which is distinct from a formal certification assessment.
Assessment under the methodology requires the provision of evidence across the seven weighted dimensions (Trust, Context, Distribution, Product, Governance, AI Integration, Autonomy). Evidence includes internal governance documents, technical documentation, deployment architecture descriptions, incident logs, and human oversight records. An organisation that is unwilling to provide evidence at the required level of specificity cannot receive a certification score above the Pre-Assessment level.
The methodology is designed for organisations deploying AI agents within or into the European Union. The methodology can be applied to agents deployed in other jurisdictions, but the certification is framed in relation to the European regulatory context. We note in any certification where an agent's deployment scope extends beyond the EU and where the methodology has been adapted accordingly.
For inclusion in our editorial analysis and framework comparisons, a standard or methodology qualifies if it is published by a recognised standards body (ISO, NIST, CEN, ETSI, BSI, or equivalent), an accredited certification body, or a supervisory authority, and addresses the safety, transparency, governance, or accountability of AI systems in a substantive and documented way.
We anticipate introducing a paid Partners tier in 2027 for verified institutional participants. The programme has not launched. The following is indicative only.
Tier 1: Reference Listing. A verified carrier, broker, consultancy, or implementation partner with documented expertise in AI certification or AI risk assessment receives a named profile page on this site, including a description of their published criteria for accepting certification as a factor in underwriting or due diligence decisions. Illustrative annual range: EUR 5,000 to EUR 25,000.
Tier 2: Sponsored Research Briefing. A commissioned briefing on the relationship between certification status and underwriting, investment, or procurement criteria, clearly labelled as sponsored content, published separately from editorial coverage. Illustrative range: EUR 25,000 to EUR 100,000 per engagement.
Tier 3: Institutional Partnership. A multi-year arrangement including data licensing for the certification registry dataset, co-branded research on methodology development, and visibility across the full Authority Stack. Pricing on application.
All paid placements will be visibly labelled as commercial. Paid partnership does not affect assessment outcomes, certification scores, methodology decisions, or editorial content. An institution that enters the paid partner programme does not receive preferential treatment in any assessment, and an institution that does not enter the programme is not excluded from editorial coverage.
A certification level awarded under the Agent Certified methodology represents an independent evaluation of an AI agent deployment against a published set of criteria. It does not constitute a government accreditation, a regulatory approval, a legal opinion, or an insurance underwriting decision. Agent Certified is not a government-designated conformity assessment body under the AI Act.
The methodology is published for transparency and to enable third parties to verify the basis of any certification outcome. Users of certification data for underwriting, investment, or procurement decisions should satisfy themselves as to the relevance of the methodology to their own assessment requirements.
If you represent a standards body, certification framework, or supervisory authority whose published framework you believe qualifies for inclusion in our editorial analysis under Section 3, please write to press@agentcertified.eu with the subject line "Editorial inclusion." Include a direct link to the primary document. Inclusion is editorial and free of charge. There is no guarantee of inclusion following a submission.
To request an assessment of your AI agent deployment, use the assessment request form at request-assessment.html.
If you represent a carrier, broker, consultancy, or investor that uses or anticipates using AI certification status as a factor in underwriting, investment, or procurement decisions, and you would like to be informed when the 2027 partner programme launches, please write to registry@agentcertified.eu. We will contact you when the consultation process opens. We do not commit to launch dates beyond 2027.
The conflicts of interest applicable to this site are set out in Section 5 of the Editorial Standards. Future Proof Intelligence is the publisher and the operator of the assessment service. This dual role is disclosed. The methodology is published in full to allow independent verification of assessment outcomes.