AIToday

Anthropic may have violated its own Responsible Scaling Policy by failing to publish risk discussions for Claude Mythos as required.

LessWrong AIApr 10, 20261 min read
Anthropic may have violated its own Responsible Scaling Policy by failing to publish risk discussions for Claude Mythos as required.

Summaries like this, in your inbox every morning.

Sign up free →

3 Key Points

  1. Anthropic's RSP 3.0 required publication of risk discussions within 30 days for certain internal model deployments, which Anthropic may have failed to do for Claude Mythos.

  2. The limited external research access program for Claude Mythos qualifies as a public deployment under RSP guidelines, which should have triggered immediate publication requirements.

  3. The violation appears clearer for the external research access deployment than for the internal deployment, though both cases suggest potential non-compliance with Anthropic's own policies.

  4. This discrepancy raises questions about whether Anthropic is adhering to the responsible scaling commitments outlined in their own RSP documentation.

Discussion

No comments yet. Be the first to share your thoughts!

Log in to join the discussion

Related Articles

Stay ahead with AI news

Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.

Get Started Free

Free · takes 30 seconds · unsubscribe anytime

1 minute a day. The AI essentials.

200+ sources · Email / LINE / Slack

Get it free →