Singapore firms face rising shadow AI without strong governance

Only 70% report having full visibility into machine identity security.

Singaporean organisations are encountering shadow AI more frequently than any other country, with 62% reporting unsanctioned AI use at least once a month, according to identity security provider Delinea.

Of those, 38% experience such incidents multiple times per month.

Despite these concerns, Singapore ranks highest globally in AI security readiness, with 52% of organisations saying they are fully prepared to secure AI tools—compared to the global average of 44%.

However, the report highlighted a potential mismatch between perceived readiness and actual practices. Whilst 97% of firms are confident in their machine identity security, only 70% report having full visibility into these identities, and just 56% have formal AI governance policies in place.

The study, which surveyed over 1,700 IT decision-makers worldwide, noted that 91% of Singaporean businesses already use generative or agentic AI in daily operations. As usage grows, concerns have emerged around specific risks.

The most commonly cited threats include AI-generated phishing and deepfakes (54%), autonomous systems with unchecked access (52%), credential theft driven by AI (47%), shadow AI (46%), and limited visibility into AI workflows (40%).

The report also found that just 63% of global organisations have acceptable use policies for AI tools, and 64% have implemented access controls for AI agents.