US Senator Bernie Sanders has renewed his push to restrict the development of advanced artificial intelligence, arguing that powerful AI systems should not be allowed to advance beyond human control while technology companies race to build increasingly capable models.
The Vermont independent criticised OpenAI chief executive Sam Altman after the technology executive warned that society could face serious cybersecurity incidents as AI capabilities improve. Sanders argued that the public should not have to accept major security risks as the price of technological progress.
The dispute has intensified a broader debate over who benefits from AI, who bears the risks and whether governments should impose stricter limits on the technology before it becomes more powerful.
Sanders and Democratic Representative Greg Casar introduced the Ban Artificial Superintelligence Act of 2026 in September. The proposed legislation would permanently prohibit the development of artificial superintelligence, temporarily pause advanced AI development until federal safety rules are established and create a new government department to oversee the technology.
The bill has not become law. It was introduced in the Senate on 23 September and referred to the Committee on Commerce, Science, and Transportation.
Bernie Sanders Hits Back at Sam Altman Over AI Cybersecurity Risks
The latest dispute followed comments by Altman about the cybersecurity challenges associated with increasingly capable AI models.
In an interview on Politico's Decoded podcast, Altman warned that the world faces a coming wave of cybersecurity problems as AI tools become more powerful. He suggested that society might have to accept some serious cyber incidents in exchange for the benefits of keeping powerful AI models open and accessible.
Sanders rejected that trade-off in a post on X on 12 October.
'No, Mr. Altman. We won't accept a severe AI cyberattack in exchange for "liberty",' Sanders wrote.
He also argued that AI developers should be subject to the law and called for a pause on advanced AI development alongside a permanent ban on artificial superintelligence.
No, Mr. Altman. We won’t accept a severe AI cyberattack in exchange for “liberty.”
— Sen. Bernie Sanders (@SenSanders) October 12, 2026
74% of the American people don’t trust Big Tech Oligarchs. And for good reason.
Sam Altman, the co-founder of OpenAI, recently said: “There is a coming tidal wave of cybersecurity problems that…
The disagreement reflects a central tension in AI policy. Developers want to make increasingly capable systems available to users, researchers and businesses, while policymakers are concerned that the same capabilities could help malicious actors conduct cyberattacks or enable other harmful activities.
Altman's remarks did not amount to an endorsement of cyberattacks. Rather, they highlighted the difficult security trade-offs that could arise as more powerful AI tools become widely available.
Sanders Wants a New Department of Artificial Intelligence
The bill goes beyond a ban on superintelligence. It would establish a Cabinet-level Department of Artificial Intelligence with responsibility for overseeing advanced AI development and enforcing the proposed restrictions.
Under the proposal, the department would establish safety requirements, review advanced models and oversee compliance with federal rules.
The legislation also seeks to restrict dangerous AI capabilities, including systems that could help develop biochemical weapons or create more advanced AI systems without human involvement.
Sanders has argued that existing oversight is insufficient to address the risks posed by increasingly capable models.
'When you are racing towards a cliff, you don't just ease up on the gas pedal. You hit the brakes,' Sanders said when announcing the legislation in September.
The proposal would represent a major shift in US AI policy, moving from a framework centred on managing risks towards prohibiting certain capabilities altogether.
However, the bill faces substantial political hurdles. It has been referred to a Senate committee, and there is no guarantee it will advance through Congress.
AI Safety Debate Puts Big Tech Under Pressure
Sanders' proposal comes as some technology executives and AI researchers warn that increasingly powerful models could create serious security risks.
Altman has spoken about the possibility of a surge in cybersecurity problems as AI systems improve. Other industry leaders have called for safeguards and greater cooperation between governments and AI developers.
Anthropic chief executive Dario Amodei has also warned about the potential risks of advanced AI, while researchers at major AI companies have raised concerns about systems that could assist with cyberattacks or other dangerous activities.
The disagreement is not simply about whether AI should be regulated. It is also about how strict the rules should be, which systems should be covered and whether restrictions should be introduced before the technology becomes more capable.
Supporters of tighter controls argue that waiting for a serious incident could leave governments struggling to respond after the damage has occurred.
Critics of sweeping restrictions, however, warn that slowing development in the US could leave the country behind international competitors and limit beneficial applications of AI.
The debate also raises questions about how governments could enforce restrictions if researchers or companies in other countries continued developing similar systems.
Sanders has argued that international cooperation would be necessary to prevent a race towards AI systems that humans cannot control.
Who Benefits From AI — and Who Bears the Risks?
Beyond the question of AI safety, Sanders has repeatedly focused on the economic consequences of the technology.
He has criticised the concentration of AI development among a small number of powerful companies, arguing that decisions about the future of the technology should not be left to corporate executives alone.
The concerns extend to employment, wealth inequality and the distribution of productivity gains. If AI allows companies to automate more tasks, the benefits could take the form of lower costs, higher profits, improved services or reduced working hours. The outcome will depend on how businesses, governments and workers respond.
The same technology could also put pressure on jobs, increase the value of specialised computing infrastructure and concentrate economic power among companies able to afford the largest AI systems.
Sanders' proposed ban would address the potential dangers of superintelligence, but it would not by itself resolve wider questions about job displacement, wages or the ownership of AI-generated wealth.
Those questions have become increasingly prominent as companies invest heavily in computing infrastructure and more powerful models.
For workers, the central concern is whether AI will complement their skills or replace some of the work they currently perform. For businesses, the challenge is to capture productivity gains while managing security, legal and operational risks.
