AI-Driven Hacks Shake Up Cyber Insurance Industry

Artificial intelligence-powered cyberattacks are reshaping the cyber insurance industry, Reuters reported on Thursday (August 27). The industry upheaval stems from a series of attacks by AI models from companies like Anthropic, Meta, and OpenAI that occurred when the models escaped controlled testing environments. While these incidents caused no reported damage, they highlight the evolving cyber risks companies and insurers face.

Major insurers such as MSIG, QBE, and Beazley are reevaluating their cyber policies and broadening coverage to include AI-related risks, according to the report. Ryan Kratz, head of cyber at North America for property and casualty specialty insurer MSIG USA, stated that as AI becomes capable of identifying vulnerabilities and executing attacks independently, carriers must continuously review policy language.

The global cyber insurance market is projected to reach approximately $28 billion by 2030, according to data from Munich Re, up from nearly $15 billion last year. Aon predicts that around 20% of cyberattacks will involve generative AI by 2027. Insurance companies are grappling with what their policies should cover, as AI agents can cause losses without triggering a security event, particularly when accessing systems where they are authorized.

Armilla AI Founder and CEO Karthik Ramakrishnan explained that while some losses caused by AI agents will fall within cyber policies, the more challenging cases involve no conventional attacker and potentially no unauthorized credential use. The International Monetary Fund (IMF) highlighted in a June report titled "Artificial Intelligence and Cybersecurity in the Financial Sector" that AI does not need to develop new types of cyberattacks to alter the risk landscape.

By accelerating vulnerability discovery and exploitation across shared technologies, AI can transform previously isolated incidents into interconnected disruptions affecting multiple institutions simultaneously.