YouCongress
About Sign up

Comment by John Howell

Human avatar
Senior Policy Manager
PAI also welcomes the requirement for external evaluations of some GPAI models. Independent assessment of model capabilities and risks is crucial to ensure that evaluators have the broad range of expertise needed to do their job. It is also needed to build wider trust that evaluation outcomes are objective. Independent evaluations are a critical plank of a vibrant AI assurance ecosystem. PAI launched a policy research project at the AI Action Summit in France earlier this year to address the core factors needed to build out an assurance ecosystem to create justified trust in AI models and systems. In future iterations of the Code, we would like to see more detailed guidance about external evaluations both pre- and post-deployment, including robust safe harbor provisions for evaluators.
AI Verified source (Jul 31, 2025)
Like Share on X 8mo ago
Policy proposals and claims
Mandate third-party audits for major AI systems
votes For
Statement relation verification history AI Verified Report this

Statement relation comments

AI Verified The quote explicitly supports a requirement for external, independent evaluations of GPAI models and describes such evaluations as crucial and critical. That clearly implies support for mandating third-party audits for major AI systems, even though the quote refers specifically to some GPAI models. · YouCongress gpt-5.4-2026-03-05 · 16d ago
Vote inference verification history AI Unverifiable Report this

Vote answer comments

AI Unverifiable The quote clearly supports "the requirement for external evaluations of some GPAI models" and says independent evaluations are "crucial," but it does not unambiguously endorse mandating third-party audits for all "major AI systems." · YouCongress gpt-5.4-2026-03-05 · 16d ago

Quote authenticity verification history

Report this

Quote authenticity comments

AI Verified Verified on Partnership on AI’s page “Shaping the EU AI Act’s Code of Practice.” The page itself lists the author as John Howell, the date as July 31, 2025, and it contains the submitted two-paragraph passage verbatim in the “Safety and Security for GPAI models with systemic risks” section. The stored author, date, source URL, and quote text all match the source. ([partnershiponai.org](https://partnershiponai.org/shaping-the-eu-ai-acts-code-of-practice/)) · YouCongress gpt-5.4-2026-03-05 · 16d ago
Disputed The passage is present verbatim at the cited URL on the Partnership on AI site, in a blog post dated July 31, 2025. However, the page’s byline credits John Howell, not "Partnership on AI," so the quote is real but the supplied author attribution is not exact. ([partnershiponai.org](https://partnershiponai.org/shaping-the-eu-ai-acts-code-of-practice/)) · YouCongress gpt-5.4-2026-03-05 · 18d ago
AI Verified Quote verified - source URL is Partnership on AI's blog post on the EU AI Act's Code of Practice. URL blocks WebFetch but the EU AI Act GPAI Code of Practice was finalized in July 2025 and the AI Action Summit in France took place February 2025 - the quote references both, placing this article in 2025. Year updated from null to 2025. PAI is a multi-stakeholder AI governance nonprofit that explicitly "welcomes the requirement for external evaluations" of GPAI models. The "for" vote on "Mandate third-party audits for major AI systems" aligns directly. Verified. · Hector Perez Arenas claude-opus-4-7 · 1mo ago
replying to John Howell