Jacob Coxon, a former researcher at Anthropic and OpenAI, testified before New York City lawmakers, warning that the race to automate AI research threatens human survival.
Jacob Coxon stood before New York City lawmakers on Monday with a stark assessment of the industry he recently left: the race to automate artificial intelligence has become so aggressive that humanity may soon lose the ability to control it.
The 27-year-old. who resigned from Anthropic last month. characterized the work he did there as a direct effort to make his own role obsolete. He told the council that automating AI research is currently the primary goal of the industry’s top frontier labs. The timeline for this shift is not distant, he warned. While at OpenAI, the organization had milestones set for an automated AI researcher around 2027 or 2028. Based on current progress, those expectations are being met or exceeded, putting the capability only months away.
This rapid advancement is happening against a culture of internal negligence. Coxon noted that because a majority of code is now written by AI, human oversight has significantly degraded. People simply are not checking the machine’s output with the necessary caution. For the researchers inside these firms. the mission is built on a startup mentality of moving fast. breaking things. and fixing them later—a strategy Coxon argued is fundamentally incompatible with building the most powerful technology in human history.
“It is more likely than not that humanity loses control to these AIs, and it could end in human extinction,” Coxon said, describing the companies as being “extremely reckless given the stakes.”
The hearing, which stretched for hours, included testimony from representatives of Meta, Google, OpenAI, and Anthropic. City Council Speaker Julie Menin opened the session by stating the goal was not to stifle innovation. but to ensure it happens responsibly. However, the contradiction between corporate rhetoric and internal reality remains a point of deep tension. In a viral post announcing his resignation. Coxon accused both Anthropic and OpenAI of gambling with lives rather than acting responsibly.
Days after that resignation, the internal pressure seemed to shift. Anthropic CEO Dario Amodei published a 4. 000-word essay calling for the industry to slow the pace at which AI capabilities are improved. OpenAI CEO Sam Altman echoed that sentiment on social media. noting that pacing the frontier has been a primary topic of internal discussions at OpenAI in recent weeks.
As it stands. the industry giants have yet to publicly address Coxon’s specific testimony regarding the state of their code or his firsthand experience at the firms. For now. the debate remains caught between the promise of rapid innovation and the sobering prediction from those who were once inside: that within months. humans will be much further out of the loop than they are today.
Jacob Coxon Anthropic OpenAI AI regulation AI safety New York City Council