Demis Hassabis, CEO of Google’s AI powerhouse DeepMind and a Nobel laureate, says there are far bigger concerns with artificial intelligence than job losses — namely, the danger of powerful AI falling into the wrong hands.

Speaking at the SXSW festival in London this week, Hassabis told CNN’s Anna Stewart that while the debate over AI taking jobs is valid, it’s the potential misuse of advanced AI systems that keeps him up at night.

“I worry more about bad actors misusing these technologies,” Hassabis said. “The real challenge is ensuring that access to powerful AI is limited to responsible people, while still enabling positive, transformative use.”

His comments come in the wake of warnings from Anthropic CEO Dario Amodei, who recently predicted that AI could wipe out up to half of entry-level white-collar jobs. But Hassabis, whose work at DeepMind includes developing human-level AI, sees the danger in broader terms — especially with artificial general intelligence (AGI), a form of AI that matches or exceeds human intelligence.

Already, there are signs of that misuse. U.S. authorities recently warned that hackers have used AI-generated voice clones to impersonate government officials. Last year, a State Department report labeled AI a potential “catastrophic” threat to national security. Deepfake pornography is also spreading rapidly online, prompting recent legislative action, such as the Take It Down Act signed into law by former President Trump to criminalize sharing nonconsensual explicit content.

Despite the risks, the race to dominate AI continues — particularly between the United States and China — often without meaningful global regulation. Earlier this year, Google quietly removed language from its AI ethics policy that once barred the use of AI in weapons and surveillance systems.

For Hassabis, the time has come for international consensus on how AI should be governed.

“Geopolitics makes that difficult right now, but eventually the need for guardrails will be undeniable,” he said. “We need to agree globally on how AI should be used for the benefit of all.”

Hassabis envisions a future where AI is deeply embedded in everyday life. He described a world of intelligent “AI agents” — software assistants that can perform daily tasks, boost productivity, and offer personal recommendations, from books to social connections. Google is already developing this vision, experimenting with AI-enhanced search tools and smart glasses powered by conversational AI.

But while AI capabilities in video generation and coding are advancing quickly, the road ahead is not without bumps. Issues such as algorithmic bias, misinformation, and hallucinated outputs have led to public blunders — including U.S. newspapers publishing AI-generated book lists filled with titles that didn’t exist.

Despite concerns over automation, Hassabis maintains that AI is more likely to transform the job market than destroy it. He believes the technology could unlock new kinds of work, boost productivity, and help rebalance economic output.

“There will be massive change — that’s certain,” Hassabis said. “But as with past revolutions like the internet, we often see better jobs emerge to replace the ones we lose. Time will tell if that trend continues.”


By WafricNews Desk.