Vendors Secretly Sending Data to AI Models

A significant majority of vendors advertising AI capabilities are not disclosing their use of third-party AI subprocessors, according to a DataGrail report. This practice means companies may be unknowingly exposing customer data to unvetted AI models, undermining data privacy agreements. The report found that 63.6% of vendors fail to disclose these subprocessors, creating substantial privacy risks and potential regulatory violations, especially as AI governance struggles to keep pace with technological advancements. This lack of transparency is particularly concerning given the increasing costs associated with data breaches and the surge in privacy-related fines.