If 2024 was the year that businesses introduced use cases for AI tools, 2025 is the year of adoption across a wide spectrum of industries. And as we predicted in last year’s Looking Ahead Guide, increased adoption of AI tools will bring increased cyber risk.

Yes, security risk will be exacerbated by attackers’ use of AI tools—but cybersecurity defensive tools are already integrating AI capabilities as well. In this sense, AI won’t change cyber risk for most organizations, but it does carry the potential for more impactful losses when they do happen. One way to mitigate this risk is by buying appropriate cyber insurance limits.

But the impact of AI on cyber risk may be felt in other areas—specifically in how large language models (LLMs) are trained. Training and customizing these models for a specific use case within your business may lead to issues around intellectual property rights for the data used to train the model. Further, employees using proprietary or sensitive business information without clearing the use through a company’s IT department can create liability issues around the use of sensitive or personal information without proper disclosure or consent.

This much is clear: AI adoption is coming quickly, and many associated risks are yet to be discovered. When such uncertainty exists, a well-brokered insurance policy can provide your organization with the courage to adopt new tools and leverage the power of AI without taking on too much risk alone.