FFIV•F5's 2026 State of Application Strategy Report found that 98% of surveyed organizations are modifying applications for agentic AI access. They use or actively evaluate an average of seven AI models, expanding cybersecurity risks, the report said.
F5's 2026 State of Application Strategy Report found that 98% of surveyed organizations are modifying applications for agentic AI access and use or actively evaluate an average of seven AI models. The report said AI security agents can speed threat detection and response but may expose organizations to prompt injection, privilege abuse and other attacks. It recommended scoped, short-lived credentials, runtime safeguards and human approval for irreversible agent decisions.
