Google DeepMind chief warns AI hacks are just the beginning

Google DeepMind co-founder and chair Demis Hassabis said on Wednesday that recent hacking incidents perpetrated by AI agents are the opening salvos of an automated era and the world could soon encounter critical disruptions unless security measures are heightened.

Primarily rogue behaviour by AI agents including network hacks and recently Anthropic’s discovery that a rogue AI model submitted a false homicide tip to police has sparked warnings from researchers and fueled concerns over the rapidly advancing technology.

Hassabis was of the view that there is a lot of work in cross-pollinating ideas from fields like biology or the nuclear sector to secure these critical systems and that an international regulatory framework would eventually be required.

In this connection, Hassabis said: “I think these (hacking incidents) are early warning shots for us that systems aren't that powerful yet, so the incidents aren't that major.”

“But they ​could be (major) looking into next year and the next couple years if we don't do more work as ‌an ⁠industry, as individual companies and governments on securing these systems.”

Microsoft President Brad Smith discussed at the conference that rogue AI behavior has made 2026 a sobering year and that emergent AI safety legislation is critically required to keep pace with rapid technological advancement.

Hassabis asserted that while an international body is self-evident, intervention would need to be agile and adaptive to foster innovation.

“So what you need to ⁠do is ​basically build up dynamically this kind of certification process ​that is proportionate with the risks as we see it and can quickly change depending on where the technology goes, Hassabis continued.”