ASL-3 activation for Claude Opus 4
Anthropic turns on ASL-3 protections for Claude Opus 4 as a precaution, its first use of that safeguard tier under its Responsible Scaling Policy.
Deployment safeguards narrowly targeted at CBRN weapons help, including Constitutional Classifiers on inputs and outputs and a bug bounty, plus security controls against weight theft such as egress bandwidth limits. Anthropic said it had not yet determined that Opus 4 definitively crossed the capability threshold.
- Date
- Thursday, 22 May 2025
- Lab
- Anthropic
- Kind
- infra
- Access
- paper only
'First ASL-3' means the first Anthropic model under ASL-3 protections. Claude Sonnet 4.5 (2025-09-29) also shipped under ASL-3; Haiku 4.5 under ASL-2.
Sources
This record was checked against its sources on 6 October 2026. How we check