AI news ·
Countries and AI Firms Back Kill Switches and Whistleblower Protection
Eighteen countries and ten AI firms agreed to support kill switches for AI systems and protections for whistleblowers.

Eighteen countries and ten AI companies, including Anthropic and OpenAI, have signed a joint statement endorsing the use of kill switches and whistleblower protections for artificial intelligence systems. The agreement establishes a policy baseline for safety measures across participating nations and firms.
The commitments focus on two concrete mechanisms: technical kill switches that can halt AI operations and legal safeguards for individuals who report risks or misconduct. While the statement sets a shared policy foundation, its practical effect will depend on how the technical controls are implemented and how enforcement is carried out by signatories.