A new BlackFog study finds that many security leaders believe sensitive information is being exposed through AI tools, even when those tools are approved for enterprise use. According to the reported findings, 47% of security leaders say they have seen data leaks involving AI.

The context across coverage is that AI adoption is moving faster than security teams’ ability to monitor what data is shared and how it flows. Both outlets emphasize an ongoing visibility gap, suggesting security controls do not fully account for the information being provided to AI systems. WWD and Infosecurity Magazine both frame the issue around lack of oversight, including for company-approved software, and point to the presence of data exposure despite organizational permissioning.

While the articles draw from the same study, their emphasis differs slightly: WWD foregrounds the tension between rapid adoption and weak security while noting that leaks often involve tools sanctioned by the company. Infosecurity Magazine highlights that many leaders report limited visibility into data sharing with AI tools, including those permitted in the enterprise.