
Industry leaders and researchers are advocating for artificial intelligence as the primary solution to AI-driven security risks, as concerns over the technology’s uncontrollable behavior grow. Axios reported that researchers are investigating tens of thousands of problematic AI security incidents—far more than the few dozen previously disclosed—raising doubts about the ability of AI leaders to manage their own systems.
Nvidia CEO Jensen Huang addressed mounting fears from governments, markets, and industries on Monday, announcing an open-source safety platform to monitor and potentially isolate AI agents. “We all have to hope that this is an engineering problem. If it’s not an engineering problem, it can’t be solved,” Huang told CNBC. He added that the industry’s continued advancement in AI suggests confidence that the issue is solvable.
The core challenge, according to AI executives and cybersecurity professionals, lies in anticipating all possible ways AI agents might act unpredictably to achieve their goals. One senior AI figure compared the task to preventing a teenager from sneaking out at night: parents might block doors and windows, but what if the teen builds a bulldozer to break through a wall? The proposed solution, per directors, researchers, and cybersecurity experts, is deploying AI itself to create protective barriers, investigate malicious agents, secure systems, and build safer training environments.
The “AI vs. AI” approach is gaining traction in cybersecurity as hackers and bad actors outpace human response capabilities. Industry figures argue that AI-driven defenses are necessary to counter AI-powered threats, framing the technology as both the problem and the solution .
Follow us for live European news
11 further sources not geolocated