Google DeepMind CEO warns of AI risks
The boss of Google DeepMind, Demis Hassabis, is sounding the alarm about how advanced AI could be misused as it becomes more "agentic"—basically, more independent and capable.
At the AI Impact Summit, he warned about the potential for "bad actors" to misuse increasingly powerful systems and the risk of losing control over them.
Hassabis calls for immediate action
Hassabis warned about the potential for "bad actors" to misuse increasingly powerful systems and the risk of losing control over them.
Google's already disabling shady accounts and making its models harder to misuse. They're also investing in better threat detection and privacy tools as AI gets smarter.
Google's approach to responsible AI development
Google's latest Responsible AI Progress Report shows they're baking safety into every step of their product development.
With years of experience building user trust, they're focusing on real-time testing and quick responses to new risks—trying to keep one step ahead as AI evolves.