A new nationwide poll released this week found that 68% of US voters support legislation introduced by Sen. Bernie Sanders (I-VT) and Rep. Greg Casar (D-TX) that would pause development of the most advanced artificial intelligence systems and ban the creation of superintelligent AI. The survey, conducted as the two lawmakers push their bill through early legislative stages, underscores growing public anxiety over the pace of AI development.
The bill, which Sanders and Casar have framed as a public-safety measure rather than an anti-technology one, would direct federal regulators to halt training runs for the largest AI models until safety standards are established, and would prohibit US companies from pursuing systems designed to exceed human-level general intelligence, often referred to as “superintelligence.”
What the Poll Found
According to the poll, support for the measure was strong across party lines, with majorities of Democrats, Republicans, and independents backing a temporary pause on frontier AI training. Pollsters noted that concern about AI job displacement, safety risks, and corporate concentration of power were the top reasons voters cited for supporting restrictions.
- 68% of respondents supported a pause on training the most powerful AI models
- A majority also backed an outright ban on pursuing artificial superintelligence
- Support was highest among voters who said they were “very concerned” about AI’s impact on jobs
The findings echo other recent surveys showing bipartisan discomfort with the unregulated growth of AI companies, even as those same companies attract record levels of investment and political attention in Washington.
The Legislative Push
Sanders has been one of the most vocal critics of the AI industry’s rapid, largely unregulated expansion. He previously called for the federal government to take a 50% public ownership stake in leading American AI firms through a proposed AI sovereign wealth fund, arguing that taxpayers should share in the profits generated by technology built partly on public research and infrastructure.
Casar, who chairs the Congressional Progressive Caucus, has echoed those concerns, framing the pause-and-ban bill as an extension of efforts to prevent a small number of companies from controlling transformative technology without oversight or accountability.
“We are not talking about banning spreadsheets or search engines. We are talking about systems that some of the very engineers building them say could pose existential risks to humanity,” Casar said in a statement accompanying the bill’s introduction.
The lawmakers have pointed to warnings from inside the AI industry itself to bolster their case. Earlier this year, an Anthropic researcher resigned publicly, citing safety concerns about the pace of model development, a departure that drew renewed attention to internal debates over AI risk management.
Industry and Political Pushback
The bill faces steep opposition from the AI industry and from lawmakers who argue that a unilateral US pause would simply cede ground to competitors in China and elsewhere. Industry trade groups have already signaled they will lobby against the measure, warning that a moratorium could freeze billions of dollars in investment and slow beneficial applications of AI in medicine, science, and manufacturing.
Some tech executives and investors have also pushed back against broader efforts to regulate or redistribute AI wealth. That tension was visible earlier this year when Sanders publicly criticized Google co-founder Sergey Brin’s $100 million campaign against a California wealth tax, arguing that tech billionaires were using their fortunes to block policies aimed at curbing inequality tied to AI-driven profits.
Congressional aides say the pause-and-ban bill is unlikely to pass in its current form given divided control of Congress and the influence of well-funded tech lobbying groups. However, supporters argue that strong public polling numbers could pressure lawmakers to at least advance narrower safety provisions, such as mandatory third-party audits or disclosure requirements for frontier model training.
A Broader Pattern of AI Backlash
The poll results arrive amid a string of local and national controversies tied to AI’s rapid rollout. In Missouri, voters recently recalled a city council member who had approved tax breaks for a new AI data center, reflecting frustration over the resource and infrastructure costs associated with AI expansion. Separately, a lawsuit filed against OpenAI by a man who says a chatbot fueled a religious delusion has renewed scrutiny of AI’s psychological effects on vulnerable users, and law enforcement use of AI-powered surveillance tools, including Flock cameras in Illinois, has drawn civil liberties concerns.
Taken together, these episodes suggest that public skepticism toward AI is not confined to abstract fears about superintelligence but extends to concrete grievances over surveillance, mental health, labor displacement, and local economic costs.
What Comes Next
Sanders and Casar are expected to continue building support for the bill through the fall, including public hearings and additional polling data intended to demonstrate voter demand for AI guardrails. Whether the measure gains real legislative traction will likely depend on how effectively its backers can counter industry arguments about national competitiveness, and whether other lawmakers are willing to back restrictions that touch one of the fastest-growing sectors of the US economy.
For now, the poll offers the clearest evidence yet that a large share of the American public views the unchecked development of advanced AI as a risk worth legislating against, even if the political path to actually doing so remains uncertain.
