More than 70 UK Members of Parliament and peers have written to Prime Minister Keir Starmer urging the government to back a new bill banning the development of artificial superintelligence (ASI). The cross-party initiative follows public warnings from AI researchers, including a high-profile resignation at Anthropic over safety risks. The proposed legislation seeks to establish a legal framework prohibiting the creation of models that surpass human capabilities across most domains.
The campaign is supported by prominent academic figures, including Berkeley professor Stuart Russell, as well as the lobby group Control AI. Supporters argue that frontier labs are developing highly autonomous systems without adequate safety guardrails, posing existential national security threats including novel biohazards and uncontrolled cyberattacks. The bill’s sponsors hope to leverage the UK’s upcoming G20 presidency to forge an international coalition around ASI restrictions.
The UK government expressed skepticism regarding a total ban, stating it does not view the bill’s measures as the correct approach, though it remains open to targeted interventions for national security risks. Critics of the bill also note that a UK-only restriction would have limited impact unless matched by the U.S. and China, where the primary frontier models are actively developed.
Why it matters
Legislative efforts to ban or strictly cap frontier AI capabilities are gaining real traction within major Western parliaments.
Diverging government responses create regulatory uncertainty for labs aiming to scale general-purpose autonomous models.
Whistleblower warnings from top AI labs are directly fueling aggressive legislative proposals and political momentum.
Source: theguardian.com
