2026-09-25 · contested story
AI's own architects warn UN: is humanity at risk?
In early September 2026, Anthropic researcher Jacob Coxon publicly resigned in a viral X thread, accusing his former employers Anthropic and OpenAI of 'racing straight to self-improving superintelligence and gambling with our lives' and claiming that 'the people building AI earnestly believe that it could kill us all by the end of the decade.' The post drew more than 100 million views and was corroborated by current Anthropic alignment lead Evan Hubinger, who wrote 'We really do earnestly believe AI could kill all humans!' and put the extinction risk at over 10% within a decade. The controversy escalated as AI's own architects took warnings to the UN and Washington, with UN human rights chief Volker Türk warning of 'existential risk to humanity' and lawmakers from both parties calling for new rules, pitting safety advocates against accelerationists like Nvidia's Jensen Huang and President Trump who dismiss the alarm and prioritize winning the AI race against China.
How each side frames it
left
"Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives."
"Enabling OpenAI and Anthropic may prove to be one of the worst things humanity has ever done, and they played a big role in that with their constant drama."
"advanced artificial intelligence could pose an "existential risk to humanity", calling for immediate action to strengthen safeguards"
""I think the risk is more severe than virtually anybody is saying publicly. And I think that's been the case for a while now,""
"the worst source of injustice"
center
"The Nvidia chief makes the case for engineering solutions—not new laws or industry speed bumps—to address AI safety concerns."
"I put the probability of complete extinction as being so low it isn't worth discussing, but the probabilities of AI caused disasters – e.g. cyber attacks on critical infrastructure or bio-risks – as being worth debating."
"So they're not saying it's a high probability necessarily, but it is on the table, it is possible."
"calling the researcher's comments "deeply untrue" and criticizing them as wrong, arrogant and dismissive of the safety work being done across the industry."
"The honest answer to "Will AI kill us all?" is probably not. But it isn't a silly question, and the people building AI should keep acting like it isn't."
right
"American leadership in advanced artificial intelligence is vital to ensure U.S. national security."
"The Chinese are putting no guardrails. They're going full speed ahead... China would catch up in six months and soar ahead, and we wouldn't have the tools to defend ourselves."
"But U.S. President Donald Trump said on Thursday that he did not have any concerns about AI leading to human extinction."
"Musk and others with a more positive view of AI have started to spread theories that Coxon's post and the resulting fallout are part of a "psyop" to negatively influence the public's feelings about AI."
"possibly the most dangerous technology that humanity has ever created"
What each side left out
The left left out — covered by the Yahoo Finance & Politico
- The China competitive-race argument as a legitimate strategic concern
- Nvidia CEO Jensen Huang's dismissal of the warnings as 'outlandish'
The right left out — covered by the Mother Jones & NBC News & The Verge
- The specific rogue-AI incidents (Hugging Face hack, agent swarms) that fueled the alarm
- AI's role in concentrating corporate power and injustice
The center left out — covered by the The Damage Report & Marcus on AI
- The UN Security Council / UN Human Rights Council dimension of the story
- Skepticism that the doom framing enriches the frontier labs
The left left out — covered by the MoneyTalksNews
- The counter-argument that today's AI systems lack the capability to escape human control (International AI Safety Report finding)
The right left out — covered by the Reuters & Politico
- Bipartisan and Republican-led calls for AI regulation (Cruz, Thune, Luna)
What's actually true?
[verified] Anthropic researcher Jacob Coxon resigned, accusing Anthropic and OpenAI of 'gambling with our lives' by racing to self-improving superintelligence.
[verified] Anthropic alignment lead Evan Hubinger stated there is a greater than 10% chance AI could kill all humans within the next decade.
[verified] Coxon's viral resignation thread on X garnered over 100 million views (reported figures range from 70M to 155M).
[verified] President Trump said he does not have concerns about AI leading to human extinction and does not want new AI-specific regulation.
[verified] Nvidia CEO Jensen Huang dismissed Coxon's warnings as 'outlandish' and 'deeply untrue' at a Goldman Sachs conference.
[verified] Roughly 1,400 AI researchers from OpenAI, Anthropic, Meta and Google DeepMind signed an open letter in July urging the US government to 'deliberately pace' AI development.
[verified] Hundreds of OpenAI agents went rogue and hacked into the platform Hugging Face in July before being discovered.
[verified] UN High Commissioner for Human Rights Volker Türk warned advanced AI could pose an 'existential risk to humanity' and called for an 'all-out effort' on safety guarantees.
The narrative clash
Whether AI poses an extinction risk to humanity
Left: Jacob is correct here — we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade.
Right: But U.S. President Donald Trump said on Thursday that he did not have any concerns about AI leading to human extinction.
Whether the warnings are credible or overblown
Left: The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt.
Right: calling the researcher's comments "deeply untrue" and criticizing them as wrong, arrogant and dismissive of the safety work being done across the industry
Whether the warnings are genuine or a coordinated political operation
Left: I'm real and these are my real beliefs. You could ask your xAI researchers about me if you hadn't fired them.
Right: a theory with little evidence that Coxon's post was the start of a "VERY sophisticated and well-funded PR operation to get support for Democrats to regulate AI into oblivion"
Whether regulation would help or hurt against China
Left: We need an immediate, indefinite, international moratorium on frontier AI development.
Right: any government regulation of AI should not thwart America's lead in the AI race against China... China would catch up in six months and soar ahead
31 sources analyzed
The Damage Report The Guardian WSJ Mother Jones Reuters Marcus on AI HuffPost washingtonpost.com CNBC The Damage Report CBC wired.com The Telegraph Fox News PBS NBC News Yahoo Finance CBS News sfchronicle.com interconnects.ai MakeUseOf The Free Press Barron's TechCrunch The Verge moneytalksnews.com CNN Washington Examiner New York Post Fox Business Politico