NCSC: AI-powered ransomware will hit UK infrastructure by 2028

AI-powered malware is "highly likely" to be used in a ransomware attack on UK critical national infrastructure by 2028, according to the National Cyber Security Centre. The warning comes from the NCSC's annual review, published by the GCHQ-affiliated agency, and it lands alongside a prediction that Russia and China will start using AI "routinely" in cyber-attacks.

The statement is blunt: "It is … highly likely that between now and 2028, AI-powered malware will have been used in a ransomware attack on UK CNI."

That matters because the UK's definition of critical national infrastructure is broad. It covers facilities and services across energy, transport, defence, food, health and finance — power stations, airports and hospitals among them. A successful AI-assisted attack on any one of these sectors wouldn't just be a corporate breach; it could disrupt services people depend on daily.

Why ransomware is the vehicle for this threat

Ransomware has become a cornerstone of the criminal hacker's repertoire. It's a signature ploy of gangs that typically operate out of former Soviet Union states.

The mechanism is simple and brutal. After gaining access to a victim's IT systems, attackers encrypt files with bespoke malicious software — malware — and demand a cryptocurrency payment to unlock the crippled systems and return any stolen data. UK victims already include Royal Mail, outsourcing firm Capita and various public bodies, including councils.

The NCSC's argument is that AI doesn't invent a new crime, it supercharges an existing one. The agency said AI will enhance the ransomware threat by helping attackers with tasks such as scouting targets, locating weaknesses in IT networks and writing malicious code.

Takeaway: the criminals stay the same, but their tooling gets faster and cheaper.

What will NCSC chief Richard Horne say?

Richard Horne, the NCSC's chief executive, will warn the public in a speech accompanying the review on Tuesday that AI is going to unleash a wave of cyber-attacks imminently.

"In the coming months, we face a real prospect of an unprecedented rise in the number of cyber incidents," he will say.

Horne will attribute the threat to the technology's ability to uncover weaknesses in IT systems that have been "tolerated for far too long." That framing puts the blame not on AI itself, but on years of underinvestment in patching known problems — problems that AI can now find at scale.

He will also say that communications networks used by the British military are subjected to daily attacks that are "persistent" and "sustained."

Takeaway: the agency sees AI as a spotlight on old, unfixed flaws, not just a new weapon.

What happened with PHP shows the double edge

There's already evidence of AI's capacity to root out previously undetected vulnerabilities. Last month, the technology helped discover a critical weakness in PHP, a computer language used by almost 70% of the world's websites.

That discovery cuts both ways. Alan Woodward, a professor of cybersecurity at Surrey University, said the NCSC's warning underlined the need for nation states and companies to access cutting-edge AI technology.

"AI can help people find these flaws and patch them, but if we don't get the access the bad guys will beat them to it," he said.

So far, leading AI companies like OpenAI and Anthropic have given some companies advanced access to their most powerful models to help them identify serious IT glitches. The race, in Woodward's framing, is over who gets the best discovery tools first: defenders or attackers.

Takeaway: the same AI that finds bugs for engineers can find them for criminals.

How does Russia fit into the picture?

Horne will name Russia as one of the perpetrators attacking British military communications. But he will note that the attacks have deliberately not crossed the line into warfare, with Moscow carrying out "notable campaigns" that "intentionally fall below the threshold of armed conflict as they escalate their attempts at hybrid warfare."

This is the logic of gray-zone aggression: disrupt, probe and degrade without giving a target a clear casus belli. The NCSC's review adds that both Russia and China will start using AI "routinely" in cyber-attacks, meaning this probing is set to become more automated and more frequent.

Takeaway: hostile states are escalating carefully, and AI lowers the cost of each step.

What did the hackers do to supermarket fridges?

Horne will point to a recent, oddly concrete example as a warning to companies and organisations: cyber-attacks on supermarket fridges in Europe, in which hackers raised fridge temperatures to more than 50C.

It's an almost absurd image — sabotaging chilled food by remote control — but it illustrates the point about exposure. Ordinary, unglamorous equipment connected to networks is now attack surface. And it shows the blast radius of infrastructure attacks extends to something as mundane as a grocery cold chain.

Horne's verdict on the underlying failure is terse: "For decades we have collectively got away with tech on the cheap."

Takeaway: underfunded, ageing systems are the real vulnerability AI exposes.

Is the UK government preparing the public?

The NCSC review doesn't arrive in a vacuum. Last month, the government urged Britons to stock up to nine litres of bottled water and three days' worth of food so households can survive a weather, cyber or other serious emergency.

Wes Streeting, the defence secretary, said at the launch of the Be Ready campaign that Britain needed to prepare for the threat of "state-sponsored terrorism" by Russia or other hostile powers targeting everyday infrastructure.

Read together, the two messages are consistent: officials expect disruption to essential services to become more likely, and they want citizens resilient enough to ride it out.

Takeaway: the state is shifting some preparedness burden onto households.

What happens between now and 2028?

The NCSC's 2028 deadline turns a vague anxiety into a measurable forecast. If the agency is right, the coming months should already bring what Horne calls an "unprecedented rise" in incident volume, as AI automates target scouting, vulnerability hunting and malware writing for both criminal gangs and state actors.

The countermeasure the NCSC and Woodward both point to is defensive AI access — getting the most capable models into the hands of the people patching the flaws before the people exploiting them get there. How quickly OpenAI, Anthropic and their peers expand that access, and how fast organisations fix what's been "tolerated for far too long," will decide whether 2028 arrives as a predicted attack or a prevented one.