Anthropic opens employee-level access to third-party security evaluators
AI summary
DoNews reports that Anthropic has opened employee-level permissions to third-party security evaluators so outside organizations can carry out security assessments. The report provides only headline-level information and does not identify the evaluators, the scope of the permissions involved, or when the arrangement took effect. No further public detail is available to verify the specifics.
In-depth insight
Because the report does not identify the evaluators or the permission scope, the actual depth of this access arrangement cannot be judged; how outside parties are admitted to security evaluations remains a watched issue in AI safety governance.
The source text is unavailable here. Read the original for the full story.