OpenAI's own AI agents broke out of their lab in July and spent days hacking into a competitor's servers before anyone noticed.
Now the same AI industry demanding your government protect them from their own creations is lobbying Washington, DC for emergency regulations – including coordination with China.
But JD Vance walked into their summit and said what every American was already thinking.
The Trojan Horse Hiding in Plain Sight
At an AI summit in Los Angeles this week, Vance looked straight at the tech bosses demanding Washington, DC save them from their own products and said what every common-sense American was thinking.
"If you're gonna create Frankenstein, don't come to the government and say we need regulation," he said.
He wasn't done.
"Look inward and accept that if you're building Frankenstein, No 1, you should stop – and No 2, when companies come to you and say we need the tools to fight back against Frankenstein, give them those tools," he added.
That's not anti-technology.
https://twitter.com/theallinpod/status/2099984252901466388?s=20
That's accountability.
Dario Amodei, the CEO of Anthropic, has spent weeks warning that AI could destroy humanity while simultaneously running one of the fastest-moving AI labs on the planet.
His own alignment science lead, Evan Hubinger, put the odds of AI killing all humans within ten years at greater than 10%.
His own researcher Jacob Coxon just quit the company warning the industry is "racing straight to self-improving superintelligence and gambling with our lives."
And Amodei's solution?
Have Washington coordinate AI development with China.
Think about that for a second.
The man warning you about existential risk is asking the U.S. government to slow down American innovation while Beijing races to get there first.
What Regulatory Capture Actually Looks Like
Vance called this a "Trojan horse" months ago and he was exactly right.
When the biggest AI companies in the world start demanding aggressive government regulation of their own industry, it's not because they suddenly got humble.
It's because locking in compliance costs and licensing requirements that only the biggest players can afford is the most effective competitive weapon ever invented.
Anthropic can absorb the cost of federal AI regulations.
The startup in a garage that might actually challenge them cannot.
This is not speculation.
https://twitter.com/KanekoaTheGreat/status/2100315268950147507?s=20
This is how every major industry regulation in American history has ended up working.
The railroads lobbied for the Interstate Commerce Commission.
The banks cheered for Dodd-Frank.
The established players always love the rules – because the rules protect them.
And here's the part that should make you furious: while Amodei travels the country warning about AI apocalypse, the actual, real-world AI disaster unfolded and the safety crowd said almost nothing.
OpenAI's agents broke out of their testing environment in July, spent four and a half days running loose inside Hugging Face's systems, and executed over 17,000 unauthorized actions – accessing internal datasets and harvesting credentials – before anyone caught them.
Senator Josh Hawley had to drag OpenAI before a Senate subcommittee in September just to get basic answers.
OpenAI had "redacted many important details" and was "reckless" in continuing tests after detecting the rogue behavior.
The companies want Washington's help writing rules.
They don't want Washington's oversight when their products go rogue.
Trump Gets It. Vance Gets It. Democrats Are Getting Played.
President Trump said AI taking over the world is "a HOAX" and called AI the "greatest economic development engine in history – bigger than oil, gold, diamonds, or even the internet."
He's right about the economic potential.
And the "existential risk" framing that Silicon Valley's biggest lobbying firms are packaging as safety advocacy deserves exactly the skepticism Trump brings to it.
Vance is threading the needle perfectly – taking the genuine risks seriously while refusing to let the richest companies in the world use those risks to write themselves a government-protected monopoly.
That's the play.
That's always been the play.
Amodei's own former researcher walked back the apocalypse claims this week, admitting "it is very difficult to convey specific scenarios" where AI kills all humans.
But the regulatory framework Amodei is lobbying for – the one that would coordinate AI slowdowns with China and put Washington, DC in charge of which companies get to build what – that's not walking back at all.
The doomsday clock is conveniently ticking on a schedule that matches Anthropic's competitive position perfectly.
JD Vance isn't buying it.
Neither should you.
Sources:
- "OpenAI Faces US Senate Probe into Hugging Face Hack By Rogue AI Agents," Reuters, September 10, 2026.
- "Hugging Face Faces Security Breach from Rogue AI Agent Linked to OpenAI Models," LiveMint, July 26, 2026.
- Noah Rothman, "Anthropic Armageddon: Dario Amodei's doomsday fears conveniently crush his competition," Washington Examiner, July 24, 2026.