On a Tuesday in September 2026, Jacob Coxon left his post as a researcher at Anthropic and stepped into the public square with a warning that has long circulated in quieter corridors: that the race to build ever more powerful artificial intelligence is outrunning humanity's ability to govern it safely. His resignation is not merely a career change but a moral statement — one that implicates both Anthropic and OpenAI in what he calls a gamble with human futures. That such a voice would emerge from within Anthropic, a company founded on the premise of cautious AI development, suggests the tensio