Anthropic· Updated
Anthropic released version 3.0 of its Responsible Scaling Policy, the framework it uses to manage catastrophic AI risks. The overhaul separates what Anthropic can do alone from what needs industry-wide coordination, and adds public safety roadmaps with external expert review.
The update introduces two new accountability mechanisms. A Frontier Safety Roadmap sets public goals across security, alignment, safeguards, and policy — goals Anthropic will grade itself on publicly. Risk Reports published every 3-6 months will detail model safety profiles and threat models, with external expert review required under certain conditions.
RSP v3 also includes a candid self-assessment: capability thresholds proved more ambiguous than expected, and government safety action has lagged behind AI advances. Both the initial Frontier Safety Roadmap and first Risk Report are already published, making these commitments publicly trackable.
We're updating our Responsible Scaling Policy to its third version. Since it came into effect in 2023, we’ve learned a lot about the RSP’s benefits and its shortcomings. This update improves the policy, reinforcing what worked and committing us to even greater transparency.
Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →
More like this
