Across the UK and US, IT and security teams are reporting increased concern about potential risks associated with the rapid adoption of artificial intelligence (AI). A report, Heimdal’s State of AI Risk Management in 2026, outlines the extent of these concerns.
The integration of AI tools such as ChatGPT and Microsoft Copilot is now widespread; ChatGPT is used within approximately 71% of business IT environments, while Microsoft Copilot is used in around 68%. While these tools have improved operational efficiency, they also introduce new security considerations that organisations are still working to address. The report notes that many organisations continue to face challenges in adapting their security approaches to the evolving AI landscape.
At the same time, AI adoption is reported to be reducing workload for IT teams. Many teams indicate that AI tools have reduced the time spent on repetitive and routine tasks by roughly a quarter of the workweek, allowing more time to be allocated to higher-priority responsibilities. However, this efficiency gain is accompanied by ongoing concerns, as organisations adjust their security frameworks to account for emerging AI-related risks.
A key concern is data leakage, particularly among teams that have implemented greater visibility into their use of AI tools. Overall, only around 40% of security teams feel their current infrastructure is adequately prepared to manage AI-related threats, suggesting a gap between adoption and readiness.
Looking ahead, improvements to security systems are expected to be important for organisations aiming to both benefit from AI and manage associated risks. Building awareness, adaptability, and resilience is increasingly seen as necessary as organisations continue to integrate AI technologies.
