Singaporean enterprises are under intense pressure to loosen identity controls to keep pace with AI, even as their security teams face growing visibility and governance gaps around human and machine identities.
The Delinea report, “Uncovering the Hidden Risks of the AI Race,” finds that 95% of organisations in Singapore push security teams to relax access controls to support AI initiatives—more than the 90% reported globally—despite widespread shortcomings in AI‑identity discovery, monitoring and privilege management.
Identity gaps in an AI‑driven environment
The report shows that AI‑driven automation is rapidly expanding the number of identities, including non‑human ones (NHIs) such as AI agents and service accounts.
In Singapore, 93% of organisations report at least one identity visibility gap, with 46% citing discovery and management of machine and AI identities as their biggest challenge. Half of respondents say these gaps are most likely to persist in AI‑related environments, nearly double the rate seen in legacy systems.
A third (37%) of Singaporean organisations say AI expansion has been one of the top drivers of increased NHI risk over the past 12 months, ahead of broader automation and cloud‑native workloads.
At the same time, 86% admit they cannot always understand why an NHI performed a privileged action, underscoring weak traceability and accountability for automated identities. Nearly six in ten (58%) lack viable alternatives to standing privileged access for AI and machine identities, leaving persistent, high‑risk permissions that could be exploited during an incident.
An AI‑security confidence paradox
Singaporean firms also exhibit a pronounced “AI security confidence paradox.” While 87% say their identity‑security posture is ready for AI‑driven automation, 47% acknowledge that governance around AI‑related identities is deficient.
Organisations are almost twice as likely to rate their ability to discover and govern identities in AI settings poorly, compared with traditional systems, and only 14% say they can always explain why an AI agent took a privileged action—below the 20% global average.
“This pressure to move fast on AI is real, but identity governance has not kept pace, which exposes enterprises to significant risk,” said Art Gilliland, CEO at Delinea.
“As AI agents multiply across enterprise environments, these identities often have the least oversight. The organisations that will succeed in the AI era will be the ones that enforce real‑time, contextual access across every human, machine, and agentic AI identity.” Art Gilliland
Delinea proposes a unified approach combining cryptographic identity, just‑in‑time (JIT) authorisation, contextual access controls and full session visibility to ensure AI‑driven automation is both secure and auditable [Delinea 2026].
