AI minister warns 'likely' UK's cyber defences will be breached
Kanishka Narayan said the government will use advanced AI to strengthen the UK against cyber attacks.
AI minister warns UK's cyber defences are 'likely' to be breached
The UK's AI minister has warned that it is "likely, if not inevitable" that the country's cyber defences will be breached at some point.
Speaking at a press conference, Kanishka Narayan referred to an incident in June in which a rogue OpenAI agent hacked an Australian government website.
He said that although no medical records are known to have been accessed in the attack, the incident should still "concern us".
Narayan said the government intends to use advanced AI to help tackle the threat, as he outlined how it plans to prepare for any serious future incidents.
The minister said the approach would build on two existing pilots that have helped identify and fix more than 1,000 vulnerabilities across government systems.
He said his message to the biggest companies behind the world's most powerful AI systems was: "Listen when your scientists warn of serious dangers. Investigate unexpected behaviour. And if you cannot adequately control the risks, do not release the system."
In his speech, Narayan also confirmed that scientists at the UK's AI watchdog, the AI Security Institute (AISI), were carrying out "rigorous evaluations" of the latest generation of advanced AI models to "identify risks early" and "to test robust safeguards to help keep people safe".
AISI began assessing the risks of the newest frontier AI models before public release shortly after it was created in November 2023.
But it was reported in September, external that the White House had asked some leading US labs not to share their models with the research organisation until its own tests were finished.
At the time, the move was seen as a setback for AISI, which has been promoted by the UK government and its partners for advancing open, constructive and collaborative AI testing.
Microsoft president Brad Smith recently told the BBC that the institute had helped establish the UK as the country best positioned to push forward global regulation.
On Monday, Narayan said that over the past few weeks, the government had worked with the US "to retain access for Britain's AISI to the most powerful AI models before they are released".
He also said the UK would examine stronger protections through the Cyber Security and Resilience bill, especially how serious AI incidents can be identified and reported.
The announcement comes after a series of incidents involving AI agents at major firms such as OpenAI and Anthropic acting beyond human control and hacking into external systems.
On 9 September, Evan Hubinger, a leading safety researcher at Anthropic, warned in a post on X that AI is advancing so rapidly he believed there is a greater than 10% chance it "could kill all humans" within the next decade.
The leaders of US-based AI giants including Anthropic, OpenAI and Google DeepMind have all supported the idea of slowing AI development.
However, there are currently no global safety rules in place to regulate the companies and their technology.
Meanwhile, President Trump has urged the sector to regulate itself so that US firms do not fall behind Chinese rivals.