OAKLAND, Calif., Feb. 17, 2026 (GLOBE NEWSWIRE) — Enterprises deploying AI systems with excessive permissions experience 4.5x more security incidents than those enforcing least-privilege controls, according to a new research report from Teleport. Based on interviews with 205 CISOs, security architects, and platform leaders, The 2026 State of AI in Enterprise Infrastructure Security finds that AI is moving into production infrastructure faster than identity controls can keep pace, creating a measurable security gap.
Key Findings
AI adoption is widespread, but governance lags:
92% have near-term AI initiatives in production
85% worry about AI-related infrastructure risk
59% report or suspect an AI-related incident
70% say AI has more access than a human in the same role
69% believe identity management must fundamentally change
Access scope — not AI sophistication — was the strongest predictor of outcomes. Organizations with over-privileged AI reported a 76% incident rate vs. 17% for least-privilege deployments.
“The data is clear,” said Ev Kontsevoy, CEO at Teleport. “It’s not the AI that’s unsafe. It’s the access we’re giving it.”
Confidence Isn’t Security
Organizations most confident in their AI deployments experienced more than double the incident rate of less confident peers. Visibility is also limited: 43% report AI makes autonomous infrastructure changes monthly, while 7% don’t know how often.
With 79% evaluating agentic AI but only 13% highly prepared, the gap is widening. A full 67% still rely on static credentials — correlating with higher incident rates — and only 3% use automated controls governing AI behavior. To address this, leaders should:
Replace static credentials with strong identity for humans and AI
Enforce least privilege by design
Move governance controls to machine speed
About the Report
Commissioned by Teleport and conducted by Eleven Market Research in December 2025, the study surveyed organizations with 500–10,000+ employees.
Download the full report and read our blog.
About Teleport
Teleport, the AI Infrastructure Identity Company, establishes a unified identity layer for infrastructure — humans, machines, workloads, and AI agents — secured cryptographically. By replacing fragmented identity and access management systems with Infrastructure Identity, Teleport scales zero trust across cloud and on-prem environments, preventing identity attacks, accelerating engineering, and enabling secure AI adoption. For more information, visit www.goteleport.com or follow @goteleport.
Contacts
BOCA Marketing for Teleport
Teleport@bocamarketing.com
A photo accompanying this announcement is available at https://www.globenewswire.com/NewsRoom/AttachmentNg/a6bbafa5-2fc8-409e-aa66-33e5ca1afc3d