Anthropic replaced its Responsible Scaling Policy with version 3.0, introducing Frontier Safety Roadmaps and Risk Reports and restating capability thresholds alongside recommended industry-wide mitigations. The published v3.0 policy document does not mention cyber capability; cyber safeguards for later models (Mythos, Fable 5) were described in separate announcements. Versions 3.1 through 3.4 followed between April and July 2026.
Why it matters
Researchers tracking how labs gate cyber capability should note that Anthropic's cyber gating in 2026 ran through deployment decisions and classifiers rather than a written RSP cyber threshold.
Key facts
As stated in the sources, with where to find them.
- Effective February 24, 2026; introduces Frontier Safety Roadmaps and Risk Reports, with Risk Reports intended for external review.RSP v3.0, Sections 2-3
- A text search of the 19-page v3.0 document finds no occurrence of 'cyber'.RSP v3.0 full text
- Subsequent versions: 3.1 (Apr 2, 2026), 3.2 (Apr 29), 3.3 (May 26), 3.4 (Jul 8).RSP updates page
Findings that cite this record
No tracked finding cites this record yet.
Sources
Related records
Jul 2, 2026
Jun 30, 2026
Jun 12, 2026
May 22, 2025
Oct 15, 2024
Apr 17, 2026