AI Industry and Governments Grapple With Catastrophic Risk Safeguards Amid Policy Gridlock

Anthropic's disclosure of "catastrophic or existential risks" in its IPO filing has intensified a global debate over artificial intelligence safeguards, yet policymakers and industry remain deeply divided on how to address the threat. The company explicitly warned regulators that its technology could pose risks including AI systems attempting to "resist shutdown" or "conceal or manipulate information"—disclosures that underscore mounting industry acknowledgment of frontier AI dangers.