Bernie Sanders Tells Meta, OpenAI and Anthropic to Stop Building AI Humans Cannot Control
The US senator sent letters to three of the most powerful AI companies warning that Congress will step in with regulation if they keep moving at their current pace.
Key points
- Senator Bernie Sanders wrote directly to the CEOs of Meta, OpenAI and Anthropic in 2025, calling on them to halt AI development.
- Sanders warned that AI capabilities have reached what he called a critical risk threshold.
- He argued the companies are losing control over the technology they are building.
- The letter threatens Senate regulation if the companies do not slow down voluntarily.
Bernie Sanders has had enough. His letters to Mark Zuckerberg, Sam Altman and Dario Amodei, first reported by The Guardian, take a blunt position: the companies are building systems too powerful for anyone, including their own engineers, to fully control. He called this a "critical risk threshold" and said it has already been crossed.
Zuckerberg's response to this kind of pressure has so far been to lean in harder. When we covered his 6,500-word AI manifesto on 10 August, we found a Meta chief calling for less government interference, not more.
What exactly is Sanders asking for?
He wants a pause. Not tweaks to a safety policy or a government review panel. A full stop.
Sanders addressed his letters to the heads of three companies behind the large language models, the AI technology that powers chatbots like ChatGPT and Claude, that have spread across workplaces, schools and phones over the past two years. His argument is simple: they've moved faster than their own ability to understand what they're building. When a technology's creators say they can't fully predict what it does, that's a reason to stop, not accelerate.
What happens if they ignore him?
Sanders threatened Senate regulation. He didn't outline a specific bill, but the warning signals that at least part of Congress is willing to set rules rather than wait for the industry to self-govern.
The US has so far taken a lighter touch than the European Union, which passed a formal AI Act in 2024 setting legal limits on high-risk AI uses. A push from Sanders, even without immediate legislative support, keeps that pressure visible and on the record.
Should ordinary people care about this?
Yes, because this technology is already making decisions about them.
AI systems now screen job applications, help doctors flag medical images and generate news summaries. If the people building those systems admit they can't fully control them, that's a problem for everyone who uses them, not just policymakers. Sanders isn't a tech insider and his letter proposes no technical fix. What it does is put some of the most powerful technology companies on notice that "move fast" is no longer a free pass.
What to watch for
None of the three companies has replied publicly. Watch for whether other senators join Sanders or distance themselves, and for any White House comment: the current administration has signalled it wants less AI regulation, not more.
If you work somewhere that uses AI tools to make decisions about people, this debate will reach your workplace. Who is legally responsible when an AI system causes harm is still largely unanswered, and that gap is exactly what Sanders is trying to force Congress to close.
