All businesses face being targeted by hackers using AI tools doubling in power every four months, Technology Secretary warns
Businesses urgently need to prepare for potential hacks by AI tools whose capability is doubling every four months, the Technology Secretary has told business leaders.
Liz Kendall said UK businesses of all sizes and in any sector were under threat from cyber attacks, not just government agencies or high-profile companies.
Her stark warning comes as the potential dangers posed by the powerful Mythos tool developed by Silicon Valley firm Anthropic become apparent.
The AI bot has been deemed too dangerous to release to the public because it can apparently outperform some humans.
Researchers say it is capable of conducting cyber attacks because it is so adept at computer programming it can find bugs in decades-old code, which it can exploit.
Anthropic, which is currently embroiled in a legal dispute with the US government after it was labelled 'a supply chain risk', said Mythos had surpassed 'all but the most skilled humans at finding and exploiting software vulnerabilities'.
Experts believe that bots like Mythos with such advanced hacking capabilities could fall into the wrong hands or be developed by powers like China.
Only a dozen companies have so far been given early access to Mythos, including Apple and Microsoft, to enable them to find and fix potential weaknesses.
Technology Secretary Liz Kendall (pictured) has said UK businesses of all sizes and in any sector were under threat from cyber attacks
The stark warning comes as the potential dangers posed by the powerful Mythos tool developed by Silicon Valley firm Anthropic become apparent. Pictured: Stock image of a computer hachie in a hoodie
In a letter to UK business leaders this week, Liz Kendall and Security Minister Dan Jarvis highlighted 'the speed at which AI capabilities are increasing and the threat they potentially pose', The Telegraph reports.
Warning businesses to be ready to defend themselves against AI-powered cyber attacks after the emergence of Mythos, Ms Kendall said the government believed the capabilities of top AI tools were doubling every four months.
'The trajectory is clear, and therefore it is vital that we are prepared for frontier AI model capabilities to rapidly increase over the next year, and plan accordingly for that outcome.
'Government action alone will not be enough. Every business in the UK has a part to play.
'Criminals will not just target government systems and critical infrastructure. They will target ordinary companies, of every size, in every sector. Attackers go where defences are weakest.'
While cyber attacks previously relied on 'a small number of highly skilled criminals', she said that was 'shifting'.
'A new generation of AI models are becoming capable of doing work that previously required rare expertise.'
Earlier this month, the Bank of England and the UK's financial watchdogs issued a warning about the dangers of the new AI to UK banks and City of London firms.
Last year, a government-run survey on cyber security breaches found an estimated 612,000 businesses and 61,000 charities were targeted across the UK.
The most high-profile – the hack of Jaguar Land Rover – became the costliest cyber attack in British history, costing the UK economy an estimated £1.9bn.
It crippled JLR for months and affected as many as 5000 organisations within its supply chain.
Other high-profile hacking casualties last year included retailer Marks and Spencer, which fell prey to a serious hack in April 2025, which disrupted its online clothing and home orders, costing it around £136 million.
The Co-op was another high-profile cyber victim, with data stolen from over 6 million customers and supplies affected to the tune of more than £200 million.
