TransUnion is advancing fraud prevention by embedding adaptive machine learning into its Device Risk solution, moving beyond static, rule-based systems. By leveraging consortium-driven data and real-time learning, the platform can better detect sophisticated fraud tactics such as device masking and identity evasion. The enhancements aim to improve detection accuracy while reducing manual effort, helping businesses stay ahead of rapidly evolving threats. As fraud becomes more complex and costly, intelligence-driven, automated security is becoming essential across the digital economy.
Digital fraud prevention is entering a more autonomous phase as TransUnion announces a significant expansion of machine-learning capabilities in its Device Risk solution. Unveiled at the Merchant Risk Council MRC 2026 conference, these enhancements arrive as global businesses report staggering fraud losses totalling 534 billion dollars. The update is designed to move beyond traditional, static, rule-based systems that have become less effective against sophisticated attackers who use virtual machines, residential proxies, and identity-masking techniques to bypass security.
The new integration utilises adaptive machine learning models that continuously learn from thousands of device signals and fraud feedback data. This intelligence is sourced from a long-standing global fraud consortium, allowing the system to proactively identify anomalies and evasion attempts that would typically appear as legitimate “new” users to legacy systems. According to internal analysis, these machine learning enhancements have demonstrated the potential to improve fraud capture by up to 50 per cent while simultaneously reducing the operational overhead required to manually maintain complex fraud rules.
Steve Yin, the Global Head of Fraud at TransUnion, noted that traditional device fingerprinting has been challenged by privacy-driven technology changes and evolving fraud tactics. He explained that the industry must meet this moment with solutions that adapt in real time and connect more signals across various browsers and applications. This approach enables the effective recognition of risky behaviour even when hardware identifiers are frequently rotated or masked by fraudsters.
The urgency of these updates is underscored by the rapid rise in account takeovers, which increased by 141 per cent between the first half of 2024 and the first half of 2025. During the same period, suspected fraud in the account creation process increased by 26 per cent. These trends indicate that attackers are increasingly focusing on the earliest stages of the digital relationship. By pairing richer device-level intelligence with adaptive learning, the platform aims to secure the entire customer journey from login to transaction.
Clint Lowry, the Vice President of Global Fraud Solutions at TransUnion, emphasised that these enhancements empower customers to operate with greater confidence. By elevating both detection accuracy and system efficiency, businesses can protect their revenue without introducing unnecessary friction into the digital customer experience. The ability to recognise returning devices across diverse client environments provides deeper insight into evolving fraud trends that no single organisation could capture on its own.
As digital fraud continues to consume an average of 7.7 per cent of annual equivalent revenue for organisations globally, the shift toward intelligence-driven prevention is becoming a baseline requirement for the financial sector. TransUnion intends to continue expanding its consortium-driven insights to stay ahead of advanced evasion tactics. This commitment to innovation ensures that as the digital economy grows, the tools used to defend it remain sufficiently sophisticated to handle the next generation of cyber threats.
What this means for the industry
- Fraud prevention is shifting from rules-based systems to adaptive, AI-driven models
- Consortium data is becoming a key advantage in detecting emerging fraud patterns
- Businesses can improve fraud capture while reducing operational complexity
- Account takeovers and early-stage fraud are rising key areas of concern
- Balancing security with seamless customer experience remains critical
Photo by Markus Winkler on Unsplash
