PeopleofSuperintelligence

Researcher

Jacob Coxon

Also known as hilbertspaess

Jacob Coxon is a British mathematician and former pretraining researcher at OpenAI and Anthropic who now warns publicly about the risks of advanced AI. A silver medallist at the 2016 International Mathematical Olympiad, he studied mathematics at Cambridge, moved to San Francisco in 2023 and joined OpenAI, where he worked on GPT-4o and was among the authors of its 2024 system card. He moved to Anthropic in May 2026 and resigned in September 2026, writing on X that the labs were "racing straight to self-improving superintelligence and gambling with our lives." Out of the AI industry since then, he argues for a slowdown in frontier model development, and he testified before the New York City Council in October 2026.

Timeline

  1. 2016–2017

    Olympiad medals for Britain

    A gifted mathematician from an early age, he makes the British team for the International Mathematical Olympiad while in high school. He wins silver in Hong Kong in 2016, when the UK team places seventh of 109 countries, and bronze in 2017. AlphaGo's 2016 defeat of Lee Se-dol shows him that machines can master domains that need human intuition and creativity.

  2. By 2023

    Mathematics at Cambridge

    He studies mathematics at Cambridge before moving into AI. After graduating, he joins London's rationalist circle and briefly works in raw materials trading. He recalls GPT-3, released in 2020, as a genuine "wow" moment.

  3. 2023–2026

    Pretraining at OpenAI

    He moves to San Francisco in 2023 and joins OpenAI's technical staff, working mainly on pretraining, the process of feeding vast amounts of text, images and code into AI models. He works on the GPT-4o model.

  4. 2024

    The GPT-4o System Card

    He is one of 417 authors of OpenAI's GPT-4o System Card, posted on arXiv in October 2024. The card describes GPT-4o, an autoregressive omni model that accepts any combination of text, audio, image and video as input and is trained end-to-end across text, vision and audio.

  5. May 2026

    A move to Anthropic

    He leaves OpenAI and joins Anthropic in May 2026 to pretrain its family of AI models, drawn in part by its reputation for safety research. Meetings with Anthropic employees around the launch of its Mythos model leave him with the impression that the company is more transparent about AI risks.

  6. September 2026

    Resigning from Anthropic

    He shifts his duties from training models to safety research, then resigns from Anthropic in September 2026 and posts on X that the labs are "racing straight to self-improving superintelligence and gambling with our lives." His posts receive 153 million views within 36 hours, and Anthropic safety researcher Evan Hubinger replies with an estimate above 10% that AI systems "kill all humans" in the next ten years.

  7. September 2026

    A PBS NewsHour interview

    In a PBS NewsHour interview in September 2026, he warns that AI research could be automated within one or two years, letting AIs make themselves smarter, and calls for international coordination. Nvidia CEO Jensen Huang answers his end-of-decade scenario in September 2026, saying there is "0% chance" that 2030 is the end of the world.

  8. Since September 2026

    Testimony at New York City Hall

    He leaves the AI industry in September 2026 and argues for a slowdown in frontier model development. In October 2026, he warns the New York City Council that on the current path humanity could lose control to AIs, possibly ending in human extinction, and he appears on The Daily Show with Jon Stewart.

  1. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.
  2. The United States employs an estimated 53% of people in our data with a known employment countryOur estimates of where AI careers are concentrated now and how countries’ shares of dated positions changed over time.
  3. 46% of living people in our data with known birth and employment countries work outside their birth countryOur estimate of cross-border careers in AI, from country of birth and the country of each current organization.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionOur estimates of where the people shaping AI studied and where their careers led, from education records in cited biographies.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Jacob Coxon yet. Send us a free-licensed photo.

  1. His AI doomsday warning shook the world. Meet the 27-year-old ex-Anthropic researcher behind it. - The Herald Business biz.heraldcorp.com. Checked 19 October 2026.
  2. BMOS/BMOC: News: IMO 2016 results (14 July 2016) bmos.ukmt.org.uk. Checked 19 October 2026.
  3. The Mathematical Association congratulate the UK IMO team in their recent success 26 July 2016 m-a.org.uk. Checked 19 October 2026.
  4. He Helped Build Powerful AI at OpenAI and Anthropic. Now He's Afraid It Could Kill Us time.com. Checked 19 October 2026.
  5. Who Is Jacob Coxon? Anthropic Researcher Quits—Warns AI Could Kill Everyone - Newsweek newsweek.com. Checked 19 October 2026.
  6. [2410.21276] GPT-4o System Card arxiv.org. Checked 19 October 2026.
  7. OpenAI and Anthropic Researchers Are Warning About AI Risks time.com. Checked 19 October 2026.
  8. AI Researcher Jacob Coxon Resigns, Warns Industry "Gambling With Our Lives" deadline.com. Checked 19 October 2026.
  9. Existential risk from artificial intelligence - Wikipedia en.wikipedia.org. Checked 19 October 2026.
  10. As AI behavior raises concerns, ex-researcher Jacob Coxon warns what may lie ahead | PBS News pbs.org. Checked 19 October 2026.
  11. 'Gambling with our lives': AI researcher quits Anthropic and leaves AI entirely over what he calls a threat to humanity finance.yahoo.com. Checked 19 October 2026.
  12. AI researcher warns 'we are racing to build and grow our own adversary' in NYC hearing cnbc.com. Checked 19 October 2026.
  13. Former Anthropic Researcher Issues AI Warning for 2027: 'It's Pretty Scary' - Newsweek newsweek.com. Checked 19 October 2026.