BREAKING
Altman: Safety First for Broad AI
Regulation as Foundation, Not Brake
Two 2026 Governance Documents
Frontier GovernanceLate May 2026
Security practices
Regulatory alignment
Federal BlueprintJune 2026
Mandatory evaluations
Oversight by CAISI
Proposed Evaluation Path
1Threshold check
2Internal + external tests
3Publish results
4License & register
Support vs Skepticism
Support
Agent risks are tangible
Verification, human override
Skepticism
Performance for compliance
Favors incumbents
From Proposal to Real Rules
AI NEWS BLITZ
OpenAI's Sam Altman says safety standards must come before spreading AI widely.