House progressives including Alexandria Ocasio-Cortez and Ro Khanna signed onto a bill banning artificial superintelligence, a move critics warn hands China the AI lead.
Rep. Greg Casar of Texas introduced the Ban Artificial Superintelligence Act this week with Sen. Bernie Sanders of Vermont, and a bloc of House Democrats quickly lined up behind it.
Washington Examiner reporting describes the measure as among the most stringent AI proposals from either party, and it arrives as Democrats argue over how hard to clamp down on the technology before the next election.
Ten House Democrats have signed on. Cosponsors include Reps. Alexandria Ocasio-Cortez of New York, Ro Khanna of California, Analilia Mejia of New Jersey, Yassamin Ansari and Adelita Grijalva of Arizona, Val Hoyle of Oregon, Stephen Lynch of Massachusetts, Chuy García of Illinois, Chris Deluzio of Pennsylvania, and Nydia Velázquez of New York.
The bill would ban superintelligence, pause advanced AI development until a federal vetting regime is stood up, and put violators at risk of a 20-year prison sentence. Ocasio-Cortez and Khanna are also described as mulling 2028 presidential runs, which puts their names on one of the left’s hardest-line AI proposals at a politically convenient moment.
Casar’s proposal does not stop at a pause. Fox News reported that Sanders and Casar want a temporary halt on advanced AI, a Cabinet-level agency to police the sector, and authority to destroy systems that fail to comply.
Penalties go further still. The package includes what critics call a corporate death penalty, stripping firms of authority to do business in the United States, plus up to 20 years behind bars for individuals, language modeled on nuclear weapons statutes.
That is a criminal regime aimed at American developers, researchers, and companies. It is not a shared global freeze. China would face no matching handcuffs under U.S. law.
"If the United States stops developing advanced AI, China is unlikely to follow. Surrendering the lead in a potentially transformative technology would carry serious economic and national security costs."
That warning came from Daniel Castro of the Information Technology and Innovation Foundation. AI expert Gary Marcus added that a permanent, unilateral ban on research into superhuman AI is too broad and would guarantee the United States falls behind.
Rep. Yassamin Ansari framed the stakes in existential terms.
"We absolutely have to slow down and regulate the advancement of dangerous AI technologies that have the potential to spiral out-of-control and threaten humanity."
Rep. Stephen Lynch pointed to recent AI-led cyberbreaches and said advanced forms of the technology are already out of control.
"We haven’t really solved that control problem yet, nor have we got the alignment problem right yet."
Khanna, ranking member of the House China Select Committee, convened a hearing this week on AI safety cooperation between the United States and China. In a statement, he said AI experts at that session underscored the need for a verifiable pause on superintelligence.
The political split inside the Democratic Party is plain. Prominent progressives are stacking onto Casar’s ban-first approach. Centrist Democrats, the same reporting notes, are looking at proposals they say would address risk without smothering innovation.
That argument will matter if Democrats retake one or both chambers in November. The party is still sorting out what “AI regulation” means when the loudest voices want prison sentences and a development freeze.
The Casar-Sanders bill is not the only Democratic push this week. The New York Post reported that 17 senators led by Sen. Chris Van Hollen of Maryland sent a letter urging President Trump to pursue a formal agreement with China’s Xi Jinping on AI safety guardrails and a possible mutual pause in frontier development.
The senators warned that an international race between the United States and China without safety guardrails poses dangerous risks to the American people, including cyberattacks and bioweapons concerns. Their letter said America leads in AI innovation and must also lead on safety and international cooperation, including work with China on a slowdown or mutual pause.
Separately, Treasury Secretary Scott Bessent and China’s Vice Premier He Lifeng announced an AI notification system so both countries can alert each other if the technology goes rogue. Trump has dismissed fears of runaway AI as a hoax, a stance that puts him sharply at odds with the progressive ban caucus and with Democrats asking him to negotiate limits with Beijing.
Sanders and Casar’s bill would outlaw recursive self-improving AI and force a development pause while a new Cabinet-level regulator is built. That is a domestic chokehold. It does not bind Chinese labs, state champions, or military researchers.
American firms are already in a live contest with Chinese developers on frontier models, compute, and applications that will shape markets and national power. A U.S.-only ban on superintelligence research, paired with felony exposure and corporate shutdown authority, would pull the most capable American players off the field.
Supporters say the technology could spiral beyond human control. Critics answer that Beijing is unlikely to match any American stop order, and that surrendering the lead would carry lasting economic and security costs. The bill’s own design makes that trade-off explicit: harsh penalties at home, no enforcement reach abroad.
Casar, Sanders, Ocasio-Cortez, Khanna, and the other cosponsors want Washington to hit the brakes hard. The record they are building is one of criminal sanctions, a new bureaucracy, and a pause that applies to the United States first.
Voters who want secure borders, working markets, and a country that keeps its technological edge should watch what happens when the same political faction that stalls domestic energy and policing decides it also wants to handcuff American AI.
