With 98% of UAE respondents reporting instances of AI tools accessing more sensitive data than required, Delinea will demonstrate how runtime authorisation can help regional enterprises govern autonomous actions and strengthen compliance.
As enterprises accelerate the deployment of agentic AI, controlling what autonomous systems can access and execute is emerging as a critical security priority. Delinea will bring this challenge to the forefront at GITEX GLOBAL 2026, demonstrating how identity security and runtime authorisation can help organisations prevent excessive access and retain control over AI-driven actions.
The company’s participation comes amid growing concerns about the gap between granting AI tools access and being able to intervene when their activity becomes risky. Delinea research found that 98% of UAE respondents had experienced confirmed or suspected instances of AI tools accessing sensitive data beyond what their assigned tasks required. Yet only 16% said they could immediately revoke an AI tool’s credentials and terminate its session.
The findings underline the need for stronger, auditable access controls as regional organisations align their security programmes with frameworks such as Saudi Arabia’s National Cybersecurity Authority Essential Cybersecurity Controls (NCA ECC) and UAE government cybersecurity standards.
“Identity security and runtime authorisation give organisations a foundation for the transition to agentic AI by empowering them to know who or what is acting, limit actions to the task, and retain the ability to intervene,” said Mortada Ayad, VP – META, Delinea.
Under the theme “From Access to Action: Governing AI at Runtime”, Delinea will demonstrate how its platform extends privileged access management and just-in-time controls across cloud-native and DevOps environments. Demonstrations will cover Secret Server, privileged access management and identity threat detection, showing how organisations can discover identities, assess their access and reduce dependence on standing credentials.
“Only 16% of UAE organisations can immediately revoke an AI tool’s credentials and terminate its session, exposing a critical gap in runtime control.” — Mortada Ayad, VP – META, Delinea
A Cedar-based policy engine enforces least-privilege controls throughout a session, determining what an identity can do and for how long. The demonstrations will also show how runtime authorisation can govern AI agents operating across databases, SSH hosts, Kubernetes clusters and cloud consoles. Task-scoped access and real-time policy enforcement can allow, block or escalate an action before execution, while credential and session controls enable teams to intervene when access needs to end.
For CISOs and IT decision-makers planning their 2027 security budgets, the focus is on balancing agent autonomy with enforceable oversight. As organisations move beyond conversational AI towards agents capable of acting across business systems, security controls must operate at the point of action—not merely when access is first granted.
Delinea will be at stand H9-C30 during GITEX GLOBAL 2026, taking place from 8–11 December at Dubai Exhibition Centre, Expo City Dubai.
