Trust is becoming harder to establish in a digital economy where credentials can be stolen, identities impersonated and interactions manipulated at scale. The consequences of failing to distinguish legitimate access from malicious activity also undermine customer confidence.

The challenge is particularly relevant across Southeast Asia. The GSMA-commissioned ASEAN Consumer Scam Report 2025, based on a survey of more than 3,000 consumers across Singapore, Malaysia, Thailand, Indonesia, the Philippines and Vietnam, found that the proportion of respondents who had ever been scammed rose from 31% to 45%. Among scam victims surveyed in 2025, 68% reported financial losses.

For business leaders, the question is no longer simply whether their organisations have strong authentication. It is whether they have enough visibility to establish trust in the devices and systems requesting access.

Why Credentials Alone Are Not Enough

Passwords and multifactor authentication remain essential, but they do not tell the whole story. Stolen credentials can still be used to access accounts, and a successful login does not necessarily mean the device or activity behind it is trustworthy. Enterprises are tasked with having to recognise and reason about millions of devices in real time.

Device identity helps close this gap. Depending on the technology, organisations can use device identifiers, device fingerprints and cryptographic keys to recognise devices and assess whether an access request presents additional risk.

For example, if someone attempts to access a corporate account using valid credentials from an unfamiliar device, the organisation can — and should have the tools to — request further verification before granting access.

When AI Agents Enter the Picture

The rise of agentic artificial intelligence (AI) introduces another dimension to identity security. Unlike assistants that primarily generate responses, AI agents can retrieve information, interact with applications and execute tasks on a user's behalf.

This matters for organisations in leading regional markets like Singapore, where recent survey findings suggest that consumers are increasingly willing to delegate responsibility to AI. Some 32% said they would likely allow AI to make decisions independently on their behalf, while 41% would likely grant an AI agent read access to their primary email to save five hours a week.

As this delegation grows, organisations must distinguish between the person authorising an action, the agent carrying it out and the device or environment through which access occurs.

Device identity can provide useful context, but it cannot establish an agent's identity or authority on its own. Businesses also need to authenticate agents, restrict their permissions and monitor their actions.

Effectively, these agents are autonomous or semi-autonomous actors that must be treated as first-class non-human identities (NHIs) interacting with the enterprise's digital infrastructure. An agent, therefore, should access only the information it needs and perform only the tasks it has been authorised to undertake.

Tackling Impersonation and Digital Fraud

As increasingly sophisticated scams look to exploit the gap between what a digital interaction appears to be and what it actually is, consumers are actively telling businesses that they are concerned.

In Singapore, survey findings indicate that 40% of respondents are concerned about increased cybersecurity risks to personal data. Another 40% cite deepfake impersonation, while 38% identify AI-generated phishing as significant risks that they are concerned about.

Whether fraudsters are stealing credentials, compromising customer accounts, or creating new accounts using stolen or synthetic identity information, the challenge is to keep cybercriminals out while maintaining seamless user experiences.

Organisations therefore need risk-based authentication, which combines real-time session scoring, behavioural biometrics, device signals and AI-driven fraud detection to identify threats continuously. This then means the organisation can reserve stronger identity verification for higher-risk interactions while minimising friction.

Security Without Compromising Privacy

Stronger identity controls must also respect privacy. In Singapore, about half of respondents said personal data used in an AI interaction must remain strictly confidential, while 84% considered government regulation protecting identity data in AI important.

These findings underscore the need to balance security with responsible data handling. Device identifiers and fingerprints can become sensitive when linked to individuals or their activities. Organisations should collect only what they need and protect that information appropriately.

The same holds true as AI agents proliferate. Every agent action should be attributable and governable. To do that, organisations need to ensure that they are governed strongly with clear identities and appropriate oversight. Delegated access controls and real-time authorisation can help prevent unauthorised actions and protect sensitive information as agents operate across applications.

The agents are already here, and they are already acting, so organisations must move now to treat every agent as a known actor, with an owner, a boundary, and a record.

Making Digital Trust a Strategic Priority

As digital interactions become more complex, identity security cannot stop at verifying credentials. Organisations need to assess the user, the device, the requested access and, increasingly, the software acting on someone's behalf.

Device identity provides an important foundation for these decisions. When combined with strong authentication and risk-based access controls, businesses arm themselves to determine when access can proceed and when additional verification is needed.

For organisations across Singapore and Southeast Asia, this is about protecting customer confidence and ensuring digital services remain trustworthy as technology evolves.

In the agentic AI era, businesses must know not just which device is connecting, but who or what is acting, under whose authority and whether that action should be allowed. Device identity is one part of building that trust.

Editorial information

Published by the Global Apex Tech Editorial Desk. Partner involvement, when applicable, is disclosed above the headline. For editorial questions or source material, contact editor@globalapextech.org.