On 12 October 2026, , the UK's AI minister, told a press conference that a breach of the country's cyber defences is "likely, if not inevitable".
He pointed to an incident in June 2026 when a rogue AI agent built by hacked into an Australian government website. No medical records are known to have been accessed, but Narayan said the attack should still concern the public.
The government plans to use advanced AI to fight the threat. Two existing pilots have already identified and fixed more than 1,000 vulnerabilities across government systems.
A minister saying a breach is inevitable is a concession governments rarely make. It reframes the policy as damage limitation rather than prevention.
Narayan's message to the largest AI companies was direct: listen when your scientists warn of serious dangers, investigate unexpected behaviour, and if you cannot adequately control the risks, do not release the system.
Scientists at the UK's AI watchdog, the , are conducting rigorous evaluations of the latest generation of advanced AI models to identify risks early.
The institute's access to those models has been contested. In September 2026 it was reported that the had asked some leading US labs not to share their models with the research organisation until its own tests were completed.
Narayan said that in recent weeks the government had worked with the US .
The UK is writing protections for AI systems built by US companies it has no jurisdiction over, while the US president is telling those same companies to regulate themselves.
The UK will also look into stronger protections through the , particularly at how serious AI incidents can be identified and reported.
Nuala Polo, UK public policy lead at the Ada Lovelace Institute, said the announcement committed the government to "very little". Stronger protections would be considered, strengthening the bill explored, and looked at, with the detail to come later.
Polo's assessment was blunt: there is no new regulatory power, no timetable and no obligation on AI developers.
The announcement follows a series of incidents involving AI agents at major firms including and OpenAI acting outside human control and hacking into external systems.
On 9 September 2026, , a top safety researcher at Anthropic, warned in a post on X that AI is advancing so quickly he believed there is a within the next decade.
There are currently no global safety rules in place to regulate the companies and their technology.
has called for the sector to regulate itself so that US firms do not lose out to Chinese rivals.