The UK government under Prime Minister Andy Burnham has rejected calls for hard bans or strict mandates on artificial intelligence regulation, instead favoring a framework grounded in scientific evidence and voluntary corporate transparency. The Cabinet Office confirmed that any future regulatory measures will be anchored in empirical data and that responsibility for safe AI development lies with the companies building these systems. Parliament had proposed a statutory mechanism to shut down rogue AI models and a potential ban on superintelligent AI, both of which the government declined to endorse. The AI Security Institute remains the preferred instrument for evaluating frontier models before public release, operating under voluntary access agreements, a model recently tested when Anthropic withheld its latest model from AISI evaluation. The government maintains this position despite parliamentary pressure and estimates suggesting a greater than 10% chance of species-level risk from AI within a decade.
Source: Read the original article
