Two Anthropic researchers have warned that advanced artificial intelligence could potentially lead to human extinction within the next decade. Jacob Coxon, who recently resigned from the company, and Evan Hubinger have publicly stated their concerns regarding the rapid progression of AI capabilities.
Following these warnings, U.S. lawmakers are intensifying calls for mandatory federal regulation of AI systems. Senators Mark Kelly, Ted Cruz, and Representative Anna Paulina Luna have advocated for stricter oversight, with some proposing special sessions and legislation focused on catastrophic risks.
OpenAI has officially called for mandatory national safety requirements, arguing that voluntary commitments are insufficient. The company proposes capability-based federal rules covering cybersecurity, independent assessments, and incident reporting for advanced models.
At the state level, California Governor Gavin Newsom signed bills SB 813 and AB 1405, establishing a framework for independent AI system assessments and third-party auditing standards. This marks the first such regulatory framework in the United States.
While the industry debate continues, President Donald Trump expressed a focus on maintaining a technological lead over China rather than prioritizing extinction concerns. Meanwhile, Anthropic stated it will continue testing models for dangerous capabilities in biology and cybersecurity.
