A new global report from Delinea has revealed an accelerating identity‑security crisis as organizations race to adopt AI. According to the study, “Uncovering the Hidden Risks of the AI Race,” nine in ten enterprises are pressuring their security teams to relax identity controls in order to fast‑track AI deployment despite widespread visibility gaps and mounting risks posed by non‑human identities (NHIs).
The research, based on a survey of more than 2,000 IT decision‑makers and analysis from Delinea Labs, highlights that AI‑driven automation is rapidly multiplying the number of identities across enterprise environments. Nearly 90% of respondents acknowledged gaps in identity visibility, with the largest deficiencies tied to machine and AI agent identities. These blind spots are occurring twice as often in AI environments compared to legacy systems, leaving organizations unable to reliably track privilege usage or identify malicious behavior.
“The pressure to move fast on AI is real, but identity governance has not kept pace.”
— Art Gilliland, CEO, Delinea
Art Gilliland, CEO of Delinea, warned that speed is creating new avenues for risk. “As AI agents multiply across enterprise environments, these identities often have the least oversight,” he said. “Organizations that succeed in the AI era will be those that enforce real‑time, contextual access across every human, machine, and agentic AI identity.”
The report found that AI expansion is now the leading driver of NHI risk, cited by 42% of organizations significantly higher than automation or cloud workload growth. Meanwhile, 80% of companies admit they cannot consistently determine why an NHI performed a privileged action, underscoring major gaps in traceability and auditability. A further 59% still rely on standing privileged access for AI agents, allowing automated identities to retain broad, persistent permissions that adversaries could exploit.
Despite these gaps, most companies remain confident in their AI readiness, creating what Delinea calls an “AI security confidence paradox.” Although 87% believe their identity‑security posture is prepared for AI automation, nearly half simultaneously acknowledge deficiencies in AI‑related governance.
Delinea says enterprises need stronger discovery, contextual access controls, and real‑time validation across humans, machines, and AI agents. Its unified platform combines cryptographic identity, just‑in‑time authorization, and full session visibility to ensure AI‑driven automation operates safely and transparently.
