Hassabis: AI Threat Research 'Needs to Be Done Urgently'
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Demis Hassabis told BBC News at the AI Impact Summit in Delhi that more research on AI threats "needs to be done urgently," calling for "smart regulation" against "the real risks" of the technology.
- Hassabis identified two main threats — AI being used by "bad actors" and losing control of systems as they become more powerful — and acknowledged his firm was "only one player in the ecosystem."
- OpenAI CEO Sam Altman also called for "urgent regulation" in a Summit speech, while Indian PM Narendra Modi said countries must work together to benefit from AI.
- The US rejected global AI governance, with White House tech adviser Michael Kratsios declaring: "AI adoption cannot lead to a brighter future if it is subject to bureaucracies and centralised control."
- Hassabis said the US and West are "slightly" ahead of China in the AI race but warned it could be "only a matter of months" before China catches up.
- UK Deputy PM David Lammy said politicians must work "hand in hand" with tech firms, adding that "security and safety must come first and it must be of benefit for the wider public."
- Delegates from more than 100 countries are attending the Summit, which ends Friday with companies and countries expected to deliver a shared statement on handling AI.
Why it matters: The summit exposes a clear fault line: industry leaders (Hassabis, Altman) and most of the 100+ attending countries want coordinated guardrails, but the US — home to the leading AI labs — is actively opposing global governance. With Hassabis warning China could close the gap in months, the regulatory vacuum may widen before any joint statement emerges Friday.
