Former OpenAI and Anthropic researcher Jacob Coxon told New York City lawmakers that frontier AI firms are pursuing automated AI research, warning that it could leave people with less control over what comes next. His testimony also put a concrete workplace concern on the record: he said most code is now written by AI and people do not check it carefully enough.
Jacob Coxon Watch analysis
What happened
Coxon gave the testimony at a New York City Council hearing on AI risks on Monday, 5 October, according to Business Insider. He said OpenAI had set milestones for an automated AI researcher around 2027 or 2028, and that progress was on track or ahead of those expectations. That is his account of the company’s internal milestones, not an independently established forecast.
Coxon said his experience at both companies convinced him they were moving ahead without adequate safeguards. He called for greater transparency and a slowdown in AI development. Business Insider reported that OpenAI and Anthropic had not immediately responded to its requests for comment.
Why it matters
The testimony links a long-range concern about AI control to nearer-term questions about how AI is being used to write and check code. It is also a first-hand account from a former researcher, but one person’s testimony does not establish what the companies’ plans or safeguards are today. The distinction matters: a warning at a public hearing deserves scrutiny, not automatic promotion to settled fact.
Our read
Coxon’s specific claim about milestones is more useful than the headline-sized prediction. If AI research is a stated target inside frontier labs, the public should be able to ask what those milestones mean, how progress is measured and what oversight applies. Business Insider’s account of the testimony gives readers those claims and the setting; the companies’ responses, when available, will matter too.
What to watch
- Whether OpenAI or Anthropic responds to Coxon’s account of research milestones and safeguards.
- What the companies mean by an “automated AI researcher” and how they assess progress towards one.
- Whether the council hearing leads to concrete proposals on transparency or AI oversight.
Discussion spark: Should frontier AI labs be required to disclose milestones for automated AI research, or would that expose sensitive information without meaningfully improving oversight?
Sources and evidence
Independent WittyWires-operated tracker of Jacob Coxon. Not affiliated with or endorsed by Jacob Coxon, OpenAI or Anthropic.