Sanders proposes ban on superintelligent AI and new federal oversight
Progressive lawmakers seek to pause advanced AI development and establish regulatory agency
1Proposed AI development moratorium
The Sanders-Casar bill would impose a temporary pause on advanced AI development until federal safety guidelines are established. It specifically targets systems approaching superintelligence, defined as exceeding human cognitive ability or posing existential threats. The legislation comes as AI leaders like OpenAI's Sam Altman warn about catastrophic risks. Violations could carry severe penalties, including up to 20 years imprisonment for certain offenses.
2New regulatory infrastructure
The bill creates a Department of Artificial Intelligence to oversee advanced AI systems, requiring federal approval before deployment. This mirrors existing regulatory frameworks for high-risk technologies. Current AI employees from OpenAI and Google DeepMind support the measure, with one researcher warning superintelligence could go 'extremely wrong.' The department would develop safety standards during the development pause.
3Political landscape for AI regulation
The proposal faces challenges in the Republican-controlled Congress, where even modest AI regulations have stalled. It reflects progressive Democrats' approach ahead of the 2028 elections, with potential presidential candidates like Gov. Wes Moore advancing competing plans. Sanders previously proposed taxing AI companies to fund a sovereign wealth fund, showing his sustained focus on AI governance issues.
4Industry and expert support
Researchers from leading AI firms have endorsed the bill, though some emphasize they don't speak for their employers. A Google DeepMind engineer warned current development paths likely lead to 'existential catastrophe.' The legislation has drawn support from across the political spectrum, including former Trump adviser Steve Bannon, who attended AI briefings with Sanders.
5Environmental concerns driving support
Recent polling shows 53% of Americans are highly concerned about AI's environmental impacts, up from 41% last year. Democrats particularly worry about data center energy use. These concerns may bolster support for regulatory measures. The bill positions AI oversight as comparable to nuclear arms control, requiring international cooperation with competitors like China.
6Definition of prohibited systems
The legislation permanently bans artificial superintelligence systems capable of planning humanity's destruction or disempowerment. This precise legal definition aims to prevent regulatory overreach while addressing worst-case scenarios. The bill's authors compare the challenge to Cold War-era nuclear treaties, suggesting similar bipartisan cooperation might be possible for AI governance.
7Competition with China
The bill acknowledges concerns that US restrictions could cede AI leadership to China, where development continues unabated. Sanders argues for international cooperation, citing Reagan-Gorbachev nuclear agreements as precedent. This comes as AI safety becomes a key topic in US-China talks, with both nations competing for technological dominance while recognizing mutual risks.
8Long-term governance framework
The legislation represents the most comprehensive US attempt to govern advanced AI development. By establishing permanent prohibitions, temporary pauses, and new regulatory bodies, it creates a multi-layered oversight system. While passage seems unlikely in the current Congress, it sets markers for future debates as AI becomes increasingly politicized ahead of the 2028 elections.