Google DeepMind's Hassabis Warns of Severe AI Safety Risks

Google DeepMind co-founder Demis Hassabis and Microsoft President Brad Smith warned at a summit that current AI-driven cyber attacks are early warning signs that could escalate into catastrophic incidents within one to two years without enhanced security and agile regulations. This warning from top tech industry leaders highlights the urgent need for governments and corporations to balance rapid innovation with robust safety protocols to prevent potential systemic disruptions driven by autonomous AI agents. Hassabis noted that current AI systems are not powerful enough to cause widespread destruction yet, but emphasized the necessity of dynamic certification mechanisms that scale with risk. Meanwhile, Microsoft President Brad Smith called for balanced, sensible AI safety legislation to ensure regulatory guardrails keep pace with technological advancement.