As artificial intelligence (AI) advances rapidly, an open letter initiated by the Future of Life Institute has sent shockwaves through the tech world. Prince Harry of the UK and his wife Meghan Markle recently joined more than 800 signatories, standing alongside pioneers such as Geoffrey Hinton and Yoshua Bengio to call for a global ban on developing artificial superintelligence (ASI) until scientists and the public reach broad consensus on its safety. The initiative highlights the urgency of AI risks and exposes deep international divisions between regulation and innovation. This article provides a comprehensive analysis of the background, the positions of various parties, and the potential implications.
Overview: a shift from "pause" to an outright "ban"
The open letter, published on October 22, quickly drew global attention. The list of signatories reads like a "cross‑sector alliance": besides high‑profile public figures such as the Duke and Duchess of Sussex, the letter includes former Irish President Mary Robinson, Virgin Group founder Richard Branson, conservative commentators Steve Bannon and Glenn Beck. Heavyweights from the AI field also feature prominently—Hinton and Bengio, both Turing Award laureates who led breakthroughs in deep learning, have in recent years warned about existential AI risks. After leaving Google in 2023, Hinton said bluntly that "AI could be more dangerous than nuclear weapons." Bengio, after receiving the 2024 Nobel Prize in Physics, has continued to push a "responsible AI" agenda.
The initiative's core demand is a "ban on ASI development," rather than the earlier calls to "pause training." The letter stresses that ASI—systems that would surpass human intelligence—could lead to "uncontrollable consequences," including mass unemployment, privacy collapse, and even human extinction. Signatories call for an international treaty to ensure the ban remains in place until a scientific consensus is formed. The Duke and Duchess' participation is especially notable: as founders of the Archewell Foundation, they have long engaged with tech ethics and previously criticized social media's harms to mental health. Harry reposted the letter on X (formerly Twitter), writing, "We cannot let AI become the next nuclear crisis." The move quickly trended, with over 100,000 related posts on X—mostly news shares but also criticisms that a ban is unrealistic.
The current call is not isolated. A 2023 "pause AI" letter drew thousands of signatures but had limited enforceability. By 2025, with iterations such as GPT‑5 and Gemini 2.0, ASI risks have moved from science fiction to a tangible concern, prompting calls for more radical measures.
Deeper background of AI safety concerns
The concept of ASI traces back to Alan Turing's idea of general intelligence, but today's focus is on "existential risk." Experts warn that ASI could self‑improve, triggering an "intelligence explosion" that outpaces human progress by millennia in hours. Hinton likens it to "a super‑smart child that may not obey." Bengio adds that ASI could amplify bias, be weaponized, or worsen economic inequality. In 2025 global AI investment exceeded one trillion dollars, yet spending on safety accounted for only about 5%, underscoring the tension between speed and security.
Public anxiety is high. A Pew Research Center survey in April 2025 found 62% of Americans worry government regulation is insufficient, and 43% think AI is more likely to be harmful. A Harvard poll showed 73% of Americans support strong regulation and 95% oppose an ASI arms race. A global KPMG report covering 47 countries and 48,000 respondents found 70% of people call for national or international regulation. These figures suggest AI is perceived less as a mere tool and more as a double‑edged sword, fueling calls for bans.
Positions: consensus and divisions
Scientists and experts: Support for a ban is substantial. Max Tegmark, founder of the Future of Life Institute, said the letter "steps out of the nerd bubble" and attracted cross‑sector celebrity support. Opponents like Ilya Sutskever of OpenAI argue a ban would "stifle innovation," advocating instead for alignment research rather than prohibition. A 2025 Google DeepMind paper warned that ASI could "permanently destroy humanity," but did not explicitly call for a ban.
Governments: Positions vary. Public support for regulation in the U.S. spans both parties, but the Biden administration's 2025 executive order emphasized "risk‑based" approaches without addressing an ASI ban. Under Republican‑led Congress efforts, tech firms have lobbied for federal uniformity to avoid state‑level fragmentation. The EU has been more assertive: its 2025 AI Act is in force and bans certain high‑risk applications, with about 48% of citizens expressing confidence in its regulatory capacity. Belgium opposed parts of the law over copyright concerns. A Pew survey shows the lowest trust in China (37%), though state‑led "next‑generation AI" programs emphasize security reviews. The International Telecommunication Union's 2025 report called for a "global governance framework," but enforcement remains challenging. Overall, governments favor gradual regulation rather than a ban, fearing falling behind in the U.S.‑China AI competition.
Industry: Tech giants largely oppose a ban. Companies such as OpenAI, Google, Meta and xAI view ASI as a "holy grail," investing tens of billions in infrastructure in 2025. OpenAI CEO Sam Altman publicly opposed "pausing AGI," arguing that "safety measures" are more realistic. Google and Microsoft signed the EU AI code of conduct, while Meta declined, calling it "overly restrictive." Elon Musk of xAI, despite warning of risks, has pushed to "accelerate ASI" to compete with China. These firms lobby against state‑level restrictions for fear of hampering innovation. Yet signatories like Richard Branson show internal industry divisions.
Analysis: feasibility of a ban and global consequences
The call is powerful but faces significant hurdles. First, ASI is hard to define: is it about parameter scale or capability thresholds? Second, enforcement mechanisms are lacking—there is no U.N.‑style treaty, and private firms could pursue covert development. Critics on X argue, "Banning AI is like banning the internet—impossible." Geopolitics exacerbates divisions: the U.S. frames a "democratic AI," the EU prioritizes human rights, and China stresses sovereign control.
Potential positive effects are clear: the campaign reshapes the narrative and pushed the 2026 G7 AI summit to discuss a possible ban framework. The Sussexes' public profile could spur youth movements similar to climate strikes. Corporate pressure may increase as public backlash grows—43% of Americans view AI as a threat. In the long term, if consensus emerges, it could lead to an "AI Geneva Convention" balancing innovation and safety.
Risks remain: a ban could trigger "regulatory arbitrage," with development outsourced to weaker jurisdictions. The Brookings Institution warned that the 2025 "AI backlash" will shape future regulation.
Conclusion: at the crossroads from appeal to action
The Sussexes' involvement marks a shift of the AI debate from elite circles into the public eye, underscoring the urgency of a "human‑first" approach. Confronting ASI's dual potential requires bridging divides: scientists must forge consensus, governments must build robust frameworks, and companies must accept responsibility. Otherwise, unchecked technological acceleration could produce irreversible catastrophe. As Tegmark put it, "This is not science fiction but our choice." Amid the AI wave at the end of 2025, this initiative could be a turning point toward controlled prosperity. In the short term, reaching consensus remains difficult, especially given U.S.‑China competition and major firms' escalating investments, so gradual progress among stakeholders is likely.