National Security Risks from AI

OpenAI and Anthropic limit access to GPT-Rosalind and Claude Mythos for 'qualified customers'.
Anthropic warns of severe national security risks from restricted AI capabilities.
Heidy Khlaaf questions lack of transparency in Anthropic's AI safety claims.
White House met with Anthropic CEO Dario Amodei on AI policy.
Shannon Vallor warns AI's 'supernatural' portrayal undermines public agency.
Anthropic partners with 40+ companies to fix AI vulnerabilities and block a Chinese group.

Security | Collaboration | AI

United States | China

Dario Amodei | Heidy Khlaaf | Shannon Vallor

4 days ago