How frontier AI is reshaping cyber security – Editorial
Artificial intelligence is changing the face of cyber security worldwide – and we’re not ready for the risks, experts say.
The recent National Cyber Security Centre (NCSC) Cyber Threat Report 2026 notes the cyber threat landscape “is changing faster than ever”.
The centre received 4673 cyber incident reports in 2025/26.
Of these, 369 incidents were serious enough to be triaged for specialist technical support because of their potential national significance, up from 331 the previous year. Nearly a quarter of these, 86, were linked to state-sponsored actors.
Four incidents were classified as “highly significant”.
On top of that, reported direct financial losses totalled $23.8 million.
The report urges business and organisational leaders to prepare for the transforming cyber landscape and the threats it brings, including “increasingly aggressive cyber criminal activity”.
But don’t forget, it’s not just leaders and their organisations who are directly affected – we could all be affected by cyber crime.
A news report this week told of families suffering delays in coronial investigations following a major cyber attack in which sensitive court information was captured. The harm trickles down.
The NCSC report says the big driver of the evolving threat is the most advanced artificial intelligence systems, known as frontier artificial intelligence (AI), making sophisticated cyber capabilities available to a much wider range of people.
In the report, NCSC deputy director-general Catriona Robinson says frontier AI models are putting tools previously available only to “advanced nation states and the largest organisations” in the hands of individuals and groups.
She says we may not know where advanced AI will ultimately take us, but one thing is clear: it will make life easier for cyber criminals. It can find security weaknesses, identify targets, crack passwords and produce increasingly convincing scams and deepfakes.
Robinson says AI offers many opportunities, but “is already being used by malicious actors to increase the speed, scale and sophistication of cyber attacks”.
She says the next generation of frontier AI models could automate attacks, identify vulnerabilities and enable “highly personalised” targeting of organisations and individuals.
And we’re not talking the distant future; the NCSC says it believes that by early next year, these “malicious actors” could have access to advanced AI capabilities.
We’ve spent years worrying, quite rightly, about AI taking our jobs, using precious resources or even helping students cheat. Now to add to those fears, we have AI lowering the expertise needed to become a sophisticated cyber attacker.
But are we ready?
As artificial intelligence rapidly reshapes the cyber landscape and frontier AI is set to supercharge both risks and opportunities, corporate awareness is crucial.
Cyber security needs to be treated by boards and chief executives in much the same way they treat financial, health-and-safety or business-continuity risks.
Our ongoing trust in and reliance on technology depends on it.