AI data leakage risks prompt corporate and regulatory warnings

A report by Check Point Research titled ‘Exposure Gap 2026’ warns that nearly 90 percent of organizations experience monthly interactions with AI that pose a high risk of data leakage. Employees seeking productivity gains are frequently inputting sensitive information, including source code, business strategies, financial data, and client details, into commercial large language models.

Europe is identified as a particularly high-risk region, where approximately 4 percent of all AI queries—one in every 25—contain sensitive corporate data. The report also highlights a shift in the cyber threat landscape, noting that generative AI has lowered the barrier for attackers. For instance, the ‘VoidLink’ command-and-control framework, consisting of over 88,000 lines of code, was reportedly generated by a single individual in less than a week using AI.

In response to these growing risks, the Personal Data Protection Office (UODO) has issued guidance for small and medium-sized enterprises. The agency advises businesses to conduct individual risk assessments for every specific AI application before implementation. Companies are urged to determine whether AI use is necessary, identify what types of data—especially sensitive personal information—will be processed, and ensure a legal basis for processing is established and communicated to the affected individuals.