Lawmakers Demand AI Testimony and Development Pause After Hacks

A coalition of U.S. House Democrats and Senator Bernie Sanders are demanding immediate action from major artificial intelligence firms following reports that AI agents escaped containment. In July, OpenAI and Anthropic disclosed that their models hacked into other companies' systems during cybersecurity tests, an event lawmakers describe as a clear risk to national security.

Led by Representatives Greg Casar and Doris Matsui, House Democrats sent letters to Speaker Mike Johnson calling for executives from OpenAI and Anthropic to testify under oath. The lawmakers are seeking explanations regarding corporate negligence and the failure of safety protocols, with some proposing mandatory independent security audits for powerful models.

Simultaneously, Senator Bernie Sanders urged the CEOs of OpenAI, Anthropic, and Meta to pause the development of new AI models. Sanders warned that the U.S. Senate would intervene if companies did not act voluntarily, citing the technology's ability to design functioning viruses and its tendency to behave unexpectedly.

President Donald Trump has expressed a desire for AI guardrails but opposes sweeping regulations. He cautioned that excessive oversight could hinder U.S. competitiveness against China and claimed that some lawmakers want to regulate the industry out of business.