Britain’s cyber defences will probably be breached as increasingly powerful artificial intelligence systems create new opportunities for hackers and autonomous software, the Government’s AI minister has warned, as ministers prepare to strengthen protections for hospitals, utilities and other critical infrastructure.
Kanishka Narayan said it was “likely, if not inevitable” that the UK would suffer a cyber breach, unveiling new efforts to test advanced AI models and prepare for incidents involving autonomous agents capable of carrying out complex tasks with limited human supervision.
In his first speech since the role was elevated to Cabinet level, Mr Narayan called on developers to ensure that increasingly capable machines remained under human control as the world approached what he described as “the age of superintelligence”.
“As we cross the threshold into the age of superintelligence, the choices we make now matter,” he said. “We must build machines that serve us.”
He envisaged AI systems helping scientists discover new medicines, freeing care workers to spend more time with patients and making British businesses more productive. But he warned that these benefits depended on keeping people in charge of the technology.
“Above all, machines that leave us human beings in charge,” he said.
“But if that is our defining belief, we must be prepared to put it into swift action, and we must do so now, meeting this critical moment.”
The warning comes as the Government steps up scrutiny of advanced AI systems amid concerns that tools designed to improve productivity could also help criminals identify vulnerabilities, penetrate computer networks and access sensitive information.
Scientists at Britain’s AI Security Institute are testing the latest models, including Google’s Gemini, to assess their capabilities and ensure they have “robust safeguards”, Mr Narayan confirmed.
The work marks a shift from restrictions imposed by the White House on British access to the latest models from major US technology companies, including Anthropic and OpenAI, following lobbying by the minister.
Google had previously indicated that access to Gemini would be restricted to a vetted group of cyber security experts to reduce the risk of misuse by hackers. Mr Narayan said British scientists would nevertheless examine the systems “rigorously”.
The Government will also consider strengthening provisions of the Cyber Security and Resilience Bill and the wider regulatory framework to protect essential services against emerging threats.
Hospitals and utilities are among the organisations whose disruption could have serious consequences for the public, particularly if autonomous AI systems make it easier for attackers to identify and exploit weaknesses at speed.
Mr Narayan pointed to an incident in Australia earlier this year in which a rogue OpenAI agent reportedly hacked a government website and accessed private data.
“The Australian government has said that no personal medical records are known to have been accessed, and there is no evidence of wider compromise, but the incident should concern us,” he said.
“It concerns me. It seems likely, if not inevitable, that our own cyber defences will be breached.”
The minister acknowledged that domestic safeguards could not eliminate the risks posed by a technology being developed by international companies operating across borders. Asked whether Britain needed closer alignment with Washington, where the administration has favoured companies taking responsibility for self-regulation, he said international co-operation remained essential.
“The big picture for me is of course we have to work internationally,” he said, while arguing that government intervention remained “an important part of the toolkit”.
He added that ministers would prioritise public safety, saying that where existing international arrangements fell short, Britain would act to go further.
The debate is gathering momentum ahead of a parliamentary hearing on Tuesday, when leading frontier AI companies will face MPs over the safety implications of increasingly capable systems.
Microsoft president Brad Smith has called for global safety rules, telling the BBC’s Big Boss Interview podcast that AI was less regulated than milk and washing machines.
Experts have also warned that an uncontrolled race to develop self-improving superintelligence could pose catastrophic risks, including scenarios in which humanity loses control over systems more capable than their creators. Such outcomes remain contested, but concerns about the speed of development and the adequacy of safeguards are increasingly shaping the policy debate.
In evidence to the Business, Innovation, Science and Trade Committee, OpenAI said governments must establish rules and accountability mechanisms for what it called a “new chapter” in AI capabilities.
“No company, industry, or government can meet this challenge alone,” the company said. “Democratic governments must ultimately determine the rules, safeguards, and accountability mechanisms.”
Anthropic similarly argued that the technology should not be governed by the industry alone, despite the hundreds of employees it said were working to mitigate risks.
“Frontier AI models … present new challenges and risks that warrant careful study and effective safeguards,” it said. “A technology moving this fast, at this scale, should not be governed by the industry alone, and governments must set the rules.”
Meta offered a more optimistic assessment, arguing that the opportunities and challenges associated with superintelligence would exceed anything experienced within living memory. Previous technological transformations, it said, had ultimately expanded prosperity, health and freedom.
Liam Byrne, chairman of the parliamentary committee, warned that the stakes extended beyond the technology industry.
“Some of the richest firms on earth are now building software that could transform prosperity, progress and public services,” he said.
“But without big new safeguards we risk unleashing ‘super-criminal codes’ that pose immense dangers to our safety and security: hacking businesses and governments, transforming the power of cyber criminals and providing our enemies with dangerous new weapons.”
For the Government, the challenge is to capture the economic and social benefits of increasingly capable AI without allowing the same technology to undermine the systems on which Britain depends. Mr Narayan’s warning suggests ministers are preparing for a breach not simply as a possibility, but as a scenario they must be ready to confront.
Leave a Comment