The warning lands as governments and industry scramble to understand how quickly artificial intelligence can change the balance between attackers and defenders.

Britain’s cyber chief said Tuesday that the UK will “likely face artificial intelligence cyberattacks on its critical infrastructure by 2028,” as criminals and state-backed actors expand their use of the technology.¹ The concern is not simply that AI will produce a new class of cyberweapon. It is that it can make existing malicious work—finding weak points, crafting convincing lures and scaling intrusion attempts—faster and easier to repeat.

The timeline is stark. By 2028, according to the warning, attacks assisted by AI could be aimed at infrastructure whose disruption would reach far beyond a single company: the systems that underpin essential public and economic life. China and Russia were specifically identified among the states increasingly using AI in cyberactivity, alongside criminal groups.¹

A separate account of the assessment framed the prospect even more bluntly: Chinese and Russian AI cyberattacks could become “routine” by 2028.² That wording underscores the central fear in the GCHQ warning—not an isolated technological shock, but the normalization of AI-enhanced pressure against critical networks.

For British officials, the message is therefore one of urgency. AI may also help defenders detect and contain threats, but the immediate warning is that adversaries are already incorporating it into their operations. The test over the next two years will be whether protection of critical infrastructure improves at least as quickly as the tools available to those trying to breach it.