Marsh says there are signs hackers are using AI tools and agents to find vulnerabilities
The head of Marsh Risk network cyber business for the Americas and Canada said a meaningful share of cyber incidents can be attributed to AI. Clients reported 105 incidents in the last two weeks of October, more than double the 48 reported in the last two weeks of May.
Meredith Schnurr, who leads Marsh Risk network cyber business for the Americas and Canada, told Reuters this week that there are signs hackers are using artificial intelligence tools and AI agents to find security vulnerabilities and that a meaningful share of cyber incidents can be attributed to AI. Clients of the firm reported 105 security incidents in the last two weeks of October, more than double the 48 recorded in the last two weeks of May.
With AI, attackers find and exploit long-standing weaknesses faster, including unpatched software vulnerabilities, exposed systems and weak authentication mechanisms. Schnurr attributed the recent surge to rapid AI tool adoption, wider deployment scale, and longer enterprise response times for containing attacks and handling incidents, arguing AI is not itself a risk but amplifies risk in every domain. As these tools become easier to obtain, attackers need fewer resources to mount larger campaigns and the pool of potential targets grows. Data firm GlobalData expects corporate demand for cyber insurance to accelerate global premium growth over the next two years, with premiums up 6.9 percent to 16.6 billion dollars in 2026 and rising 10 percent to 18.3 billion dollars in 2027.