As the US government embraces the term “Super Intelligence,” a very different movement is asking the world to stop building it. The Statement on Superintelligence, released in October 2025, calls for a prohibition on developing superintelligent AI until it can be shown to be safe and has public support. Here is what it says, who signed it, and why it matters now.
What the statement says
The statement is a single sentence:
“We call for a prohibition on the development of superintelligence, not lifted before there is broad scientific consensus that it will be done safely and controllably, and strong public buy-in.”
It was published on October 22, 2025 by the Future of Life Institute (FLI), a nonprofit that works on risks from advanced technology.
The two conditions
The prohibition would stay in place until both conditions are met:
- Broad scientific consensus that superintelligence can be built safely and controllably.
- Strong public buy-in: genuine democratic support for building it.
That makes it a conditional ban, not a permanent one. The signers are not saying superintelligence must never exist; they’re saying it shouldn’t be built until we know how to control it and the public agrees.
Who signed it
The list stood out for its range. It brought together people who rarely agree on anything:
- AI pioneers: Geoffrey Hinton and Yoshua Bengio, two of the researchers most credited with modern deep learning.
- Tech founders: Apple co-founder Steve Wozniak and Virgin’s Richard Branson.
- Public figures: former Irish president Mary Robinson, and Prince Harry and Meghan Markle.
- Political voices from the right: including former Trump adviser Steve Bannon.
It launched with around 800 notable signatories, according to CNBC, and passed 133,000 signatures by January 2026.
How it differs from the 2023 “pause” letter
In March 2023, FLI organized a widely discussed open letter calling for a six-month pause on training AI systems more powerful than GPT-4. The 2025 statement is different in two ways:
| 2023 open letter | 2025 Statement on Superintelligence | |
|---|---|---|
| Target | Training the largest AI models | Developing superintelligence specifically |
| Length | Six-month pause | Until safety and public-support conditions are met |
| Form | Detailed letter | One-sentence statement |
Importantly, the statement does not call for banning today’s AI tools. Its target is superintelligence: systems far beyond human ability across the board.
Why it matters now
In September 2026, President Trump renamed AI “Super Intelligence” for the federal government and told the UN he would encourage the technology rather than rein it in. (Read our explainer on the order.) Elon Musk then announced his AI company would become “SpaceXSI.”
That creates a sharp contrast. The same word, superintelligence, is now being used:
- by the US government and parts of industry as a positive brand for AI, and
- by safety researchers and statement signers to describe something that should not be built yet.
Expect that tension to shape debates over AI regulation, international agreements and public opinion through 2027.
The arguments for and against
Supporters say:
- A system much smarter than humans could be impossible to control once built.
- There is currently no scientific consensus on how to align such a system with human values.
- Decisions this large should not be made by a few companies without public input.
Critics say:
- A prohibition could hand the lead to rival countries or companies that ignore it.
- “Superintelligence” is hard to define, which makes a ban hard to enforce.
- The benefits, from medicine to science, could be enormous, and delays have costs too.
What happens next
The statement is not law; it is a public call. Its influence will depend on whether governments, international bodies or the AI labs themselves adopt any of its conditions. In the UK, the House of Lords Library has already published a briefing on whether superintelligent AI development should be stopped, a sign the idea has reached legislators.
The bottom line
The Statement on Superintelligence asks for a halt on building superintelligent AI until it’s proven safe and the public agrees. Backed by AI pioneers and an unusually broad coalition, it stands in direct contrast to the US government’s new pro-“SI” stance, and it defines one side of the biggest debate in technology.
Learn more: What is super intelligence? · AI 2027 explained
Enjoyed this? Get the SI Briefing
One short email a week: what changed on the road to super intelligence, and why it matters. No spam, unsubscribe anytime.
Frequently asked questions
What is the Statement on Superintelligence?
It is a one-sentence public statement released by the Future of Life Institute on October 22, 2025, calling for a prohibition on developing superintelligence until there is broad scientific consensus that it can be done safely and controllably, and strong public buy-in.
Who signed the superintelligence statement?
Signers include AI pioneers Geoffrey Hinton and Yoshua Bengio, Apple co-founder Steve Wozniak, Richard Branson, former Irish president Mary Robinson, Prince Harry and Meghan Markle, and Steve Bannon, along with more than 133,000 people by January 2026.
Is the statement a ban on AI?
No. It targets superintelligence, meaning AI far beyond human ability in nearly every field, not today’s AI tools. And it is conditional: the prohibition would lift once safety and public support are established.
How is it different from the 2023 AI pause letter?
The 2023 open letter asked for a six-month pause on training very large AI systems. The 2025 statement asks for an open-ended prohibition specifically on superintelligence, lasting until scientific and public consensus conditions are met.
Sources
- Axios: Hinton, Bannon, Bengio sign superintelligence pause (Oct 22, 2025)
- CNBC: Wozniak, Branson and hundreds urge superintelligence ban (Oct 22, 2025)
- Wikipedia: Superintelligence ban
- House of Lords Library: Superintelligent AI, should its development be stopped?
- DDG: Toward a conditional ban on superintelligence development
- Scientific American: Trump rejects AI regulation in U.N. address