Ex-Anthropic Researcher's Doomsday Thread Ignites AI Backlash

Ex-Anthropic Researcher's Doomsday Thread Ignites AI Backlash

A figure seated at a desk, one hand raised mid-speech, facing the viewer.
A former Anthropic researcher's viral warning about AI risks has drawn scrutiny from President Trump and conservative commentators.Illustration: The Frank

The News

A warning from a former Anthropic researcher that artificial intelligence could soon destroy humanity has blown the AI "doomsday" argument out of tech circles and into a full public fight.

The Hill reports that an X thread posted last week by Jacob Coxon, a 27-year-old former Anthropic researcher, struck a chord that years of similar warnings from researchers never did, setting off what it calls a firestorm of debate and confusion over how fast the technology is being built.

The reaction has split along familiar lines: safety researchers saying the risk is now real, and critics — including President Trump — saying the panic is overblown.

Timeline

Last week

Coxon publishes the X thread warning that AI could soon destroy humanity. The Hill reports it broke past the usual tech audience and drove the debate mainstream.

Sept. 12, 2026

RealClearPolitics publishes a commentary piece, "On Controlling AI: Humanity Cannot Outsource Its Survival," arguing the question of controlling the technology cannot be handed off.

Sept. 13, 2026

RealClearPolitics runs a second piece, "The Madness of the Anti-AI Crowd," pushing back on the alarm.

Sept. 13, 2026

The Hill reports on the viral debate. The same day, FRANCE 24 airs a segment on the race to make AI safer, noting that Trump has dismissed some of the concerns as exaggerated.

Reactions

On FRANCE 24, AI philosophy professor Raphaël Millière said a real-world AI security incident this summer showed the risks are no longer hypothetical — while stressing that it does not mean "AI will kill us all." His argument: the priority is coordinated action by companies and governments to manage safety risks that are increasingly real, not apocalypse talk.

Trump has dismissed some of the doomsday concerns as exaggerated, according to FRANCE 24.

On the right, RealClearPolitics has run commentary on both sides of the question in two days — one piece arguing humanity cannot outsource its survival to the companies building the systems, the other dismissing the anti-AI crowd outright.

AI leaders themselves, FRANCE 24 reported, have warned that the technology is advancing faster than the ability to control it.

What's Next

The immediate question is whether the argument stays online or turns into policy. The safety camp's stated ask — per Millière — is coordinated action by companies and governments, which means rules that do not yet exist.

Expect the fight over who gets to define the risk, and who gets to regulate it, to keep running while the labs keep shipping.

More

What makes this round different is not the warning. Researchers have been raising alarms about AI's potential danger for years, The Hill notes. What is different is who is listening: a single thread from a 27-year-old ex-employee of one of the leading labs did what years of papers and panels did not.

Poll

How seriously should Washington take AI "doomsday" warnings?

Very — regulate now
0.0%
Somewhat — watch it closely
0.0%
Not at all — it's hype
0.0%

Trump Doubles Down on United Ireland Amid Protests

Kosovo Parliament Re-elects Kurti After Long Deadlock

Boris Johnson Escapes Russian Drone Strike in Ukraine

Johnson Wants AI CEOs Summoned Before Any Law

Saudi Pipeline Outage Threatens 4% of Oil Supply

Xi Presses BRICS to Broker Middle East Peace

Russia Hits Ukraine Power Sites Before Winter

Birds, Not Drones, Scramble NATO Jet Over Lithuania

Scotland, Wales, NI Leaders Demand Breakup Referendums

Beijing Bans Drone Possession Inside the City

Modi Warns BRICS of Supply Chain Shocks

American Pope Marks 9/11 With Peace Appeal

Turkey Blocks LGBT Accounts, Detains 26 in Raids

Israeli Strike Kills Two in Gaza City

Iran Says Ship Struck Near Hormuz, One Dead

Anthropic CEO Urges AI Slowdown; Altman, Musk Agree

Labour Eyes Donation Curbs After Reform's £72m Haul

Badenoch Leads UK Backlash Over Trump's Ireland Remark

Russian Missile Destroys Ukrainian Tennis Star's Odesa Home

GOP Governor Candidate Swims From Sinking Plane