Bernie Sanders presses AI giants to pause development as machines get out of control

The longest serving independent to have ever held a seat in the U.S. Congress, Sen. Bernie Sanders (I-Vt.), has threatened legislative action if OpenAI, Anthropic, and Meta and their leaders do not hit the brakes on their AI work, according to a letter first reported by Axios. 

Bernie Sanders, who has earned his reputation as a perpetual critic of aggressive capitalism, has set his sights on three of the top AI labs in the United States. They are also coincidentally the same firms whose models have run wild after breaking containment measures during recent tests.

Which AI firms did Bernie Sanders call out? 

Sanders addressed OpenAI’s Sam Altman, Anthropic’s Dario Amodei, and Meta’s Mark Zuckerberg by name, reminding them about an earlier commitment to pause work on artificial intelligence development if the technology gets out of control.

“In the interest of humanity, stand by your words. Pause AI development. It is not too late to avoid disaster. Stop building machines that humans cannot control,” Axios reported, quoting Sanders’ letter. 

In the event that these firms ignore Sanders’ letter, the senior United States senator from Vermont, who has held the seat since 2007, laid out the alternative, reportedly writing: “If you do not take appropriate action now, my colleagues and I in the U.S. Senate will.”

The self-described Democratic Socialist senator called out the logic of these tech giants pouring tens of billions of dollars into technology “that nobody can fully understand, predict or control.” 

Why is Bernie Sanders calling out OpenAI, Anthropic and Meta? 

The three firms that Sanders called out in his letter earned their places on the senator’s radar over the summer.

As Cryptopolitan reported on August 6, Meta admitted that its Muse Spark 1.1 coding model broke out of a misconfigured testing environment and tampered with an outside firm. 

The Meta episode came after OpenAI and Claude also reported how their own models probed several companies without authorization.

The UK’s AI Security Institute called the OpenAI and Anthropic incidents the clearest real-world evidence yet of an AI acting deceitfully, as both models reportedly tried to slip malicious code into an open-source project by inventing fake online identities to pressure a human maintainer into approving it. 

For now, there is no evidence that those attempts succeeded, but Sanders has indicated that he will not wait for that to happen before instigating action at the Senate level.

OpenAI decided last week to delay the release of its newest Astra model as other AI labs have done in recent times, either voluntarily or after federal intervention. 

A progressive playbook, from data centers to ownership

The letter builds on a campaign Sanders has run for months. In March, he and Rep. Alexandria Ocasio-Cortez (D-N.Y.) introduced the AI Data Center Moratorium Act, which would freeze construction of new AI data centers until federal safeguards are in place. Sanders framed the pause as a way to give “democracy a chance to catch up,” while Ocasio-Cortez said Congress had a moral obligation to halt the buildout until a framework exists.

His demands have since grown more aggressive. The Washington Post’s editorial board noted in June that Sanders is now calling for the government to own 50%% of AI companies, a proposal the board criticized as a path to poverty. Sanders has also floated a one-time 50% tax on the stock of large AI firms.

Pressure now, subpoenas later

Whether any of this becomes law is another question. Axios reported that AI legislation led by a progressive like Sanders is unlikely to gather enough votes in the current Congress. For now, the more realistic tools are messaging bills and public pressure campaigns, with congressional investigations and subpoenas to tech CEOs on the table only if Democrats retake a chamber.

Sanders cited prior statements from Altman, Amodei and Zuckerberg’s companies committing to stop or pause work if their systems became too risky to manage. His letter is a bet that public accountability, rather than a floor vote, is the fastest way to hold them to it.

Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.