<span>Story: :: Google DeepMind CEO Demis Hassabis warns of dangers as AI-assisted research enters ‘golden age’</span><span>:: New Delhi, India / February 18, 2026</span><span>“I think we’re on the cusp of an absolutely incredible transformation that will bring incredible benefits to science and medicine, which I’m particularly passionate about.”</span><span>:: February 17, 2026</span><span>“As these systems become more powerful, I think, you know, there are at least two risks that we’ve been concerned about. One is bad actors, human actors… individuals, but also potentially nation-states using these systems and repurposing them for harmful purposes because they serve a dual purpose… But as we get closer to AGI (artificial intelligence), I think we may be moving into a kind of agency era, we should call it, of systems being more autonomous, and I think we’ll see a lot of that happen this year and next year.”</span><span>“I think we will soon need to worry about things like biological and cyber risks. I think the current system is already pretty good at cyber, and I think we need to make sure cyber defenses are stronger than the attack vectors.”</span><span>Speaking at the AI Impact Summit in New Delhi, Hassabis also sounded the alarm, highlighting urgent concerns about the potential misuse of AI, particularly in the fields of biotechnology and cybersecurity. </span><span>Indian officials have positioned the India AI Impact Summit, which opened on Monday (February 16), as a platform to amplify the voices of developing countries in global AI governance. Delhi marks the first time this global event has been held in a developing country. </span>
