Anthropic CEO Dario Amodei has proposed that frontier AI labs should commit to embedding independent safety evaluators within their organizations. In a blog post on Saturday, Amodei said these embedded evaluators would check whether a company is actually following the training, deployment, operational, and safeguards practices it claims to follow. The evaluators would have employee-like access, including desks in the offices, access badges, and company laptops, to verify safety practices and report incidents.
Related Articles
Don't miss out on breaking stories and in-depth articles.