Anthropic bringing in Accenture moves AI safety testing toward an audit industry

Anthropic bringing in Accenture for AI safety testing is a sign that frontier-lab oversight is starting to professionalize. The Financial Times reports that Dario Amodei wants labs to embed third-party testers more deeply, which shifts safety from internal claims toward outside review.

That matters because the frontier model business is asking customers, governments, and investors to trust systems that are difficult to independently inspect. If outside testers can get enough access, safety evaluation could start to look more like financial audit, cybersecurity assurance, or aviation investigation.

The risk is shallow certification. Third-party testing only matters if evaluators have real access, technical independence, and the ability to publish uncomfortable findings rather than rubber-stamp a release.