Unofficial, untracked use of AI tools known as shadow AI is spreading across everyday work inside financial institutions, Finextra reported. The issue is less AI itself than organisations not knowing what data staff enter, where and how often. Employees use browser-based tools without procurement or system links, and some bypass restrictions by using personal devices. The activity can evade existing controls and leave little audit trail, raising questions over reliance on policies and blocking.

Zero Networks has launched an AI Segmentation feature designed to control enterprise AI agents based on identity and block AI-enabled lateral movement in networks, SiliconANGLE reported. The company said rising use of AI agents is exposing security gaps, including lack of controls over access to sensitive systems, faster lateral movement driven by AI attacks and compliance operations that cannot keep pace with change. The feature includes lateral-movement controls, agent monitoring and a compliance and risk engine.