Jacob Coxon, who spent three years doing pretraining research at both OpenAI and Anthropic, resigned from Anthropic in early September 2026, declaring on social media that neither company was acting responsibly as they raced toward self-improving superintelligence. In a Fox News appearance with Sean Hannity, Coxon pressed for a formal agreement between the United States and China on AI development, arguing that without coordination the technology's ability to improve itself could spiral out of human control before anyone had a chance to respond. His post drew sweeping attention online and was echoed by Anthropic's own alignment science lead, who separately acknowledged the industry lacks a proven plan for safely managing superintelligent systems. Entrepreneur Andrew Yang was among the public figures who urged Americans to take Coxon's warning seriously.

Coxon's sudden and public exit has not gone unquestioned. His media blitz arrived amid reports of intense "crunch time" pressure inside frontier AI labs, prompting some observers to ask why his account carried so much more weight than earlier, similar warnings from other departing researchers. Skeptics have also pointed to the financial stakes tied to leaving a fast-growing AI company, noting that resigning researchers often walk away from unvested equity. Even so, Coxon's central argument, that AI is unlike past technological scares because it can improve itself faster than regulators or companies can react, has resonated across the political spectrum and kept the debate over AI safety squarely in the national spotlight.