Former Anthropic researcher Jacob Coxon will testify at a New York City Council hearing on artificial intelligence safety, where he is expected to call the company’s practices “extremely reckless.” The hearing, convened by Council Speaker Julie Menin, will examine kill switches and regulation for advanced AI systems.

Coxon has warned that AI firms cannot fully control the models they are building. He has said that a majority of the code at leading AI labs is now written by AI itself, raising questions about oversight. His testimony is expected to push for mandatory safety brakes on powerful AI systems.

The hearing reflects growing political pressure on AI companies as their systems become more capable. Lawmakers in several countries are weighing rules on testing, transparency, and emergency shutdowns. Anthropic has said it takes safety seriously and follows responsible scaling policies. Coxon’s appearance is likely to add fuel to the debate over how fast AI development should be allowed to move.