Continued use of shadow AI tools puts companies at risk

Netskope's report showed that the widespread use of unsanctioned AI tools continues to pose serious security risks for companies, Cybersecurity Dive reports.

The analysis found that 47% of people using generative AI rely on personal accounts that fall outside company oversight, based on data collected from October 2024 to October 2025. While the figure is down sharply from 78% the year before, the security firm said unmanaged AI use still creates security gaps that attackers could exploit. Use of company-approved AI accounts rose from 25% to 62%, but the share of users switching between personal and enterprise accounts increased from 4% to 9%. Netskope said this indicates that organizations may not yet provide the convenience or features employees want.

The report warned that personal AI use can lead to compliance issues, unsecured connections between AI services and internal systems, and data exposure. Netskope said incidents involving sensitive data sent to AI tools doubled year over year, with companies averaging 223 incidents per month. The firm recommended stronger AI governance policies.