Sanders and Casar unveil bill to permanently ban superintelligence and create a Department of AI

Start with what the bill actually does, because the weight of this story sits in the text. According to the AP, the legislation would permanently prohibit artificial superintelligence, which it defines as a system that exceeds human cognitive ability or has sufficient capabilities to plan and execute the destruction or disempowerment of humanity. It would also temporarily pause development of the most advanced AI systems until new federal safety rules are established by the newly created Department of Artificial Intelligence. After that, advanced AI systems would need federal approval before they could be deployed. The penalties for violations are reported to be substantial, including up to 20 years in prison in some cases. The AP reports that a statement shared with the news agency showed current employees at AI companies and experts signing on in support of the bill.

The second thing worth noting is who backs it, because that is what makes this more than political noise. Per the AP, Juan Felipe Ceron Uribe, a researcher in OpenAI's Safety Systems, said in a statement supporting the bill that superintelligence could either go extremely right or extremely wrong, and that we should not be playing such games. Swante Scholz, a software engineer at Google DeepMind who said he was not speaking on behalf of his employer, said that on the current path of development the most likely outcome is an existential catastrophe for humanity, and that a ban on superintelligence development would be a positive change for the foreseeable future. Both are current employees rather than former staff turned commentators, which gives the endorsement list more weight than a typical advocacy letter.

The third angle is timing, and the backdrop says more than the bill text. The AP reports Congress has so far done little to rein in the AI industry even as some of its most prominent leaders warn about potentially catastrophic risks. It notes that President Trump has resisted recent calls for greater government oversight, telling the United Nations General Assembly on Tuesday that artificial intelligence will be renamed super intelligence. Sanders told the AP that Trump has very little understanding of what this issue is about and that he is mainly concerned about the economic implications of a slowdown. Meanwhile the debate over AI regulation is scrambling political alliances in Washington: the AP reports Maryland Democratic Gov. Wes Moore announced a plan to regulate AI on Tuesday, while California Democratic Gov. Gavin Newsom signed an executive order last week to accelerate implementation of a California law that calls for independent oversight of AI companies.

The fourth angle is where the bill sits politically, and how likely it is to pass. The AP reports the Sanders and Casar bill faces long odds in the Republican-controlled Congress, where lawmakers have struggled to coalesce around even less sweeping AI regulations. At the same time, the report says it offers a window into how the progressive wing of the Democratic Party is approaching AI regulation ahead of November's midterms and a 2028 presidential campaign; Casar chairs the Congressional Progressive Caucus. The report also notes potential 2028 presidential candidates have rushed in recent weeks to release proposals showing they take the issue seriously. As for Sanders himself: the report describes the two-time presidential candidate, now 85, as having focused especially on the issue in recent months, introducing legislation in June to create a sovereign wealth fund financed through a one-time tax on the stock of the largest AI companies, hosting colleagues for a briefing with experts last week and attending a conference focused on the issue alongside Republicans including former White House adviser Steve Bannon.

The fifth thread is the polling and the China variable, which together determine how far calls to slow down can travel. The AP reports a recent poll from the AP-NORC Center for Public Affairs Research and the Energy Policy Institute at the University of Chicago found about half of Americans, 53%, are extremely or very concerned about artificial intelligence's environmental impacts, up from 41% last year; Democrats drive much of the increase, but the survey also reveals broader worries about the local impact of data centers. The report adds that AI safety is expected to be a central part of talks between Trump and China's Xi Jinping at the White House this week, and that the prospect of slowing development in the United States while China continues has been one of the central arguments against sweeping restrictions. Sanders' response is that avoiding the most dangerous forms of AI would ultimately require cooperation between the two countries, comparing the challenge to nuclear arms control during the Cold War and citing Ronald Reagan, an arch-conservative, as someone smart enough to sit down with Gorbachev and work out a nuclear treaty.

Two things remain to watch. First, the hardest mechanism in the bill is not the ban itself but the requirement for federal approval before deployment: the ban defines systems that do not yet exist, while the approval gate governs products already running, and the latter is more likely to change the industry's calendar. Which agency applies it and against what standard are both still unanswered. Second, the composition of the endorsements matters: current safety staff at frontier model companies publicly signing on to a bill that would restrict their own employers' core work is rare in other industries, and it reflects how openly divided the field has become about the pace of development. Every fact and quotation here comes from the Associated Press report, with no speculation added.