Congress is now debating a law to permanently ban a technology that does not exist yet, and more than 700 public figures already signed a letter demanding it happen.
Quick Take
- The Future of Life Institute wants a ban on superintelligent AI until scientists agree it is safe and the public agrees to allow it.
- Senator Bernie Sanders and Representative Greg Casar introduced the Ban Artificial Superintelligence Act in Congress.
- Over 700 signers, including Nobel Prize winners and celebrities, backed the push for a ban.
- Critics say the ban targets a system with no clear definition and could push research underground.
What The Ban Would Actually Do
The Future of Life Institute’s “Statement on Superintelligence,” released in September 2025, calls for stopping development of AI smarter than humans. The group says the ban should stay in place until two things happen: broad scientific agreement that such systems can be built safely, and strong public support for building them at all.
That is not a vague warning. It sets a specific bar. Nothing moves forward without both scientific consensus and public buy-in. The Institute has repeated this same language across its official pages, making clear this is not a one-time press release but a fixed organizational position.
In September 2026, that advocacy turned into an actual bill. Sanders and Casar introduced legislation described in reporting as mandating “a complete halt in superintelligent AI development,” permanently stopping any system built to surpass human intelligence. The bill also includes a temporary pause on advanced AI systems until a federal regulator sets up safety rules, a two-track approach pairing a permanent ban with a short-term freeze.
Who Signed On, And Why It Matters
The open letter behind this push pulled in an unusually wide crowd. Reporting counted more than 700 signers, including Nobel laureates, tech industry veterans, sitting policymakers, and public figures like Prince Harry and Meghan Markle. That range gives the campaign reach most policy letters never get. It also opens the door for critics to call the coalition scattered rather than scientifically unified.
Representative Casar defended the urgency in a television interview, arguing lawmakers “still have time to put these rules in effect” before superintelligent systems get built. Politico also reported the push follows a string of AI-powered cyberattacks, tying the bill to real, recent incidents rather than pure speculation. That connection to actual events gives supporters a stronger hook than abstract fear alone.
The Case Against A Ban
Not everyone views a ban as the answer. The Information Technology and Innovation Foundation argued the proposals “rest on speculation, not evidence” and said banning advanced AI “would not make the United States safer,” warning it would instead slow scientific progress and weaken economic and national security. Their preferred fix is oversight, not prohibition, through incident monitoring and audits of high-risk systems.
Cognitive scientist Gary Marcus, no fan of unchecked AI risk himself, publicly opposed the Sanders-Casar bill. He argued that safety should come from developers proving their systems are controlled to an independent authority, not from a blanket freeze on all advanced development. His stance shows the ban debate does not split cleanly between AI worriers and AI cheerleaders.
Wikipedia’s summary of the wider debate notes a practical problem baked into any ban: there is no precise line separating advanced AGI from superintelligence, making enforcement genuinely hard. Critics also warn that heavy restriction could just push research into unregulated, underground channels instead of stopping it. That is not a talking point to dismiss. It is the same lesson history teaches about every prohibition effort, from alcohol to encryption.
Why This Fight Is Just Getting Started
The bill has not passed a committee or become law. It remains a proposal, backed by a petition and a growing chorus of names, not an enacted regime. That does not make it meaningless. It means the real test is ahead, in hearings, amendments, and lobbying battles where tech companies will fight hard against any hard stop on their roadmaps.
For readers skeptical of both blank-check tech power and heavy-handed government mandates, the honest read is this: nobody has proven a superintelligent system already exists or has escaped human control. But nobody has proven current safeguards would hold if one did. That gap between fear and proof is exactly where Congress is now being asked to legislate.
Sources:
time.com, nextgov.com, fedscoop.com, superintelligence-statement.org, ddg.fr, securityboulevard.com, futureoflife.org, en.wikipedia.org
