Loading activity…
@prpaskov
frontier AI auditing, evals & policy | @averiorg @uniofoxford @randcorporation | views my own
AI
A post thanks SecureBio for its “leadership, voice, and work”; another account reposted the message.
AI
A participant says they spoke with NPR about third-party evaluations and praises Jingnan Huo’s article for its nuance.
AI
A post argues that separately drafted university curricula need at least one full-time role to coordinate the work and keep it aligned with emerging auditing frameworks.
AI
One of the paper’s authors announced the spotlight acceptance on September 25, 2026.
AI
A post recommends a piece by Sean McGregor about investigating the root causes of an AI incident.
AI
“You heard it here first,” a user wrote while sharing a Business Insider link about embedded evaluators.
AI
A user calls for more attention to the group’s work, highlighting a standard covering embedded audits and members including METR, Transluce, RAND, AVERI, Princeton and SecureBio.
AI
The post says strict nondisclosure agreements keep many evaluators from discussing their work publicly, even as they contribute to AI system and model documentation.
AI
Amodei says Anthropic will give third-party evaluators permanent, employee-level access to its systems to verify adherence to safety measures, report incidents and assess models’ alignment during training.
AI
The proposal targets risks between AI agents—including conflict and collusion—and calls for independent evaluators to examine systems across developers.
AI
Citing the METR/Redwood investigation, the post urges more frequent, deeper and longer voluntary audits—with more embedded experts and less reactive forensics.
AI
Benton announces his move from managing Anthropic's Scalable Oversight team to the new role.
AI
Applications opened for the 12-week Winter 2027 fellowship that includes a new auditing and governance stream.
AI
Newsletter examines safe testing environments as emerging limit on AI progress after recent pause.
