The short version
- New legislation seeks to permanently ban artificial superintelligence and temporarily halt advanced AI development pending federal safety guidelines.
- The bill proposes creating a Department of Artificial Intelligence to oversee deployment, with violations potentially carrying prison sentences of up to two decades.
- While the measure faces significant hurdles in a Republican-controlled Congress, it reflects growing internal pressure from tech employees and shifting political alliances on technology oversight.
Two prominent progressive lawmakers have introduced sweeping legislation aimed at halting the development of artificial superintelligence and establishing a new federal agency dedicated to overseeing advanced artificial intelligence systems. Senator Bernie Sanders of Vermont and Representative Greg Casar of Texas unveiled the bill, which represents one of the most aggressive regulatory proposals to emerge from Congress regarding emerging technology. The measure comes as industry leaders increasingly urge lawmakers to impose guardrails on rapidly advancing tools, even as political consensus on specific regulations remains elusive.
The proposed law defines artificial superintelligence as any system that surpasses human cognitive capabilities or possesses sufficient power to plan and execute actions that could destroy or disempower humanity. Under the legislation, such systems would be permanently prohibited. Additionally, the bill mandates a temporary pause on the development of other advanced AI models until the newly created Department of Artificial Intelligence establishes comprehensive safety rules. Once these guidelines are in place, any deployment of advanced systems would require explicit federal approval.
Enforcement mechanisms within the bill are notably strict. Violations of the ban or the development pause could result in severe penalties, including prison terms of up to twenty years for certain offenses. This punitive approach underscores the lawmakers' view that unchecked acceleration poses unacceptable risks. The creation of a dedicated department signals an intent to centralize oversight rather than relying on fragmented regulatory efforts across existing agencies, which have so far done little to rein in the industry despite warnings from prominent tech figures.
Support for the legislation extends beyond Capitol Hill into the ranks of major technology companies. Several employees at leading AI firms have endorsed the bill, expressing concern over the trajectory of current development paths. Juan Felipe Cerón Uribe, a researcher within OpenAI’s Safety Systems division, stated that superintelligence outcomes could be either extremely positive or catastrophic, arguing that society should not gamble with such high stakes. Similarly, Swante Scholz, a software engineer at Google DeepMind, warned that the most likely outcome of the current course is an existential catastrophe for humanity, framing the ban as a necessary protective measure.
Despite this internal industry support, the bill faces steep political headwinds. The Republican-controlled Congress has struggled to agree on even modest AI regulations, making the passage of such a sweeping prohibition unlikely in the near term. President Donald Trump has resisted calls for greater government oversight, placing him at odds with tech executives like Dario Amodei of Anthropic, Sam Altman of OpenAI, and Elon Musk of xAI, who have voiced concerns about potential risks. This disconnect highlights a complex political landscape where traditional party lines on technology regulation are becoming increasingly scrambled.
The introduction of this bill also serves as a strategic move ahead of the November midterm elections and the 2028 presidential campaign. Artificial intelligence is expected to become a central issue in future electoral contests, prompting potential candidates to articulate their positions. Maryland Governor Wes Moore recently announced a plan to regulate AI, while California Governor Gavin Newsom signed an executive order to accelerate independent oversight of AI companies. Sanders, who has focused heavily on this issue in recent months, previously introduced legislation for a sovereign wealth fund financed by taxes on large AI firms and hosted bipartisan briefings with experts.
Public sentiment appears to be shifting toward greater caution, particularly regarding the environmental footprint of artificial intelligence infrastructure. Recent polling data indicates that 53% of Americans are now extremely or very concerned about the environmental impacts of AI, a significant increase from 41% just one year prior. Democrats have driven much of this rise in concern, but worries about local impacts, such as the proliferation of energy-intensive data centers, are broadening across the political spectrum. This growing public anxiety provides a backdrop for legislative efforts that prioritize safety and environmental sustainability over unchecked innovation.
As the debate continues, the Sanders-Casar proposal offers a clear window into how the progressive wing of the Democratic Party intends to approach technology governance. While the immediate legislative prospects are dim, the bill forces a conversation about the limits of artificial intelligence development. It challenges both industry leaders and policymakers to consider whether current trajectories align with long-term human safety. The coming months will likely see further fragmentation in political alliances as voters and lawmakers grapple with the dual pressures of technological advancement and the need for robust regulatory frameworks.
Sources behind this briefing
Go to the original reporting
- PBS NewsHour↗Sen. Bernie Sanders unveils bill to ban artificial superintelligence and create Department of AI
