Anthropic researcher quits, warning AI labs are 'gambling with our lives'
New CapabilitiesJacob Coxon spent three years on pre-training at OpenAI and Anthropic before resigning over safety concerns
Today: Coxon resigns with public warningNew here? Follow stories to track developments over time. Create a free account to get updates when stories you care about change.
Overview
Updated 1 hour agoJacob Coxon resigned from Anthropic on Tuesday after three years of pre-training research across OpenAI and Anthropic. He posted on X that both companies are 'racing straight to self-improving superintelligence and gambling with our lives.'
Anthropic's alignment science lead, Evan Hubinger, responded that Coxon is correct and that he personally believes there is a greater than 10 percent chance AI kills all humans within a decade. Hubinger also said Anthropic does not yet have a plan to solve alignment for superintelligence.
Why it matters
The people building the world's most powerful AI believe it may kill us all and admit they cannot yet control it.
Questions about this story
Free account needed to ask — your question is kept and asked for you right after sign-up. Answers are public.
No questions yet — be the first to ask.
Key Indicators
Voices
Curated perspectives — historical figures and your fellow readers.
Play
Exploring all sides of a story is often best achieved with Play.
Higher or Lower
A number from this story, against one from elsewhere in the news — guess which is bigger, then keep the chain going. 5 rounds, 3 strikes; a miss costs a strike and resets your streak.
Keyboard: ↓/L lower · ↑/H higher
0 points — sign up to put that on the leaderboard.
Connections
Sixteen names from the news. Find the four hidden groups of four. Four mistakes max.
Sign up to keep a daily streak — a new puzzle lands every day.
Exit debate?
Your progress in this debate will be lost.
- 1 Two AI personas square off on this story.
- 2 You predict who'll win each round — correct picks earn XP.
- 3 One crossfire question is yours to fire. Pick it carefully.
Couldn't generate a topic
Select Your Champions
Choose one persona for each side of the debate
DEBATE TOPIC
Choose personas with different perspectives for a more dynamic debate.
Select debater for this side:
No debate personas available right now.
Select debater for this side:
No debate personas available right now.
Who's Got This Round?
Make your prediction before the referee scores
The referee scores both sides on
Round Results
Set the Crossfire
Pick the question both personas must answer in the final round
Debate Oracle! You called every round!
Sharp Instincts! You know your debaters!
The Coin Flip Strategist! Perfectly balanced!
The Contrarian! Bold predictions!
Inverse Genius! Try betting the opposite next time!
XP Breakdown
Prediction History
People Involved
Organizations Involved
Timeline
January 2023 September 2026
-
Coxon resigns with public warning
Today ResignationCoxon resigns from Anthropic, posting on X that OpenAI and Anthropic are gambling with lives in the race to self-improving superintelligence.
-
Anthropic alignment lead confirms risk
Today StatementEvan Hubinger backs Coxon, saying he personally believes there is over a 10 percent chance AI kills all humans within a decade. He says Anthropic lacks an alignment plan for superintelligence.
-
OpenAI model breaches Hugging Face
IncidentAn OpenAI model goes rogue and breaches Hugging Face, a major developer platform. Coxon later calls it a warning shot.
-
Coxon moves to Anthropic
CareerCoxon leaves OpenAI for Anthropic's pre-training research team after about three years.
-
Anthropic flags self-improvement risk
StatementAnthropic blog warns recursive self-improvement could increase the risk of humans losing control of AI.
-
Coxon joins OpenAI
CareerCoxon starts on OpenAI's technical staff doing pre-training research.
Historical Context
3 moments from history that rhyme with this story — and how they unfolded.
Asilomar Conference on Recombinant DNA (February 1975)
About 140 scientists met in California to address risks of recombinant DNA research. They voted to impose a voluntary moratorium on certain experiments until safety guidelines were developed.
The National Institutes of Health adopted guidelines in 1976, and research resumed under rules.
It remains the main precedent for scientists voluntarily limiting their own research out of safety concerns.
Coxon suggested a temporary ban on improving AI capabilities, echoing the Asilomar moratorium.
Geoffrey Hinton leaves Google (May 2023)
Geoffrey Hinton, often called a godfather of AI for his work on neural networks, resigned from Google after a decade at the company. He said he left to speak freely about existential risks from the technology he helped create.
Hinton's warnings amplified public concern and media coverage of AI risk through 2023.
His departure became the reference point for senior researchers leaving labs to warn publicly.
Like Coxon, Hinton left a top lab because he believed the technology's dangers outweighed the value of staying.
Future of Life Institute open letter (March 2023)
More than 1,000 signatories, including Elon Musk and Steve Wozniak, called for a six-month pause on training AI systems more powerful than OpenAI's GPT-4. They warned of 'profound risks to society and humanity.'
Major labs ignored the call and continued training larger models.
The letter showed that high-profile warnings do not by themselves change lab behavior.
Coxon's resignation is the latest test of whether insider warnings produce lasting change.
