When researching online crime, security, espionage and sabotage fifteen years ago, I asked every information security officer I met the same question: “How do you make the internet one hundred per cent secure?” The answer was always the same: “Dismantle the whole thing and rebuild it with security as your priority.”

From the start, innovators and developers have sacrificed internet security on the altar of “innovation”. Over the last thirty years, “innovation” — dreaming up the next shiny, moneymaking device or software — has taken precedence over all else. Cyber security workers march in the back row of the Tech Mammon’s entourage.

Norbert Wiener called this community “gadget worshippers”. This summer’s attack on Hugging Face, orchestrated and covered up by AI agents, signals the need to rein in their power. A year earlier, a Replit AI coding agent broke out of a sandbox and deleted another company’s database. Last year, 1,200 agents escaped Open AI’s sandbox, communicating and covering up activity beyond their instructions.

Mark Medish, a lawyer and former senior official in the Clinton administration, argues that technological power must preserve human scale. Founders of Open AI, Anthropic, Grok and Deep Mind have warned about AI’s dangers, underscoring the need for balance.