*By Luanna Vargas Shirozono
2025 was a tough year for those working in the fight against fraud, not because of any recent issues, but because fraudsters evolved very quickly and prevention structures are composed of many parts; it's not easy to be on the front lines. Conversely, fraudsters have no requirements or certifications to operate.
In a market landscape increasingly marked by fragmented journeys, complex integrations that combine legacy and new technologies simultaneously, with little end-to-end visibility, widen the gaps. When this happens, the impact is not only financial, but also diminishes trust, strains relationships, and diverts energy that should be dedicated to innovation and expansion.
It was indeed a year marked by social engineering in fraud, corporate pressure cookers, and the need for shock treatments. Perhaps the first shock of 2025 was realizing that fraud was not limited to isolated incidents; crime had become industrialized, and the market faced organized models of identity theft, balance manipulation, receipt falsification, the creation of shell companies, and the massive use of generative artificial intelligence to circumvent visual verifications. The automation that was supposed to protect was also used against the system.
Furthermore, another shock is the pressure for accelerated and often disorganized digitization, which opens up significant vulnerabilities where solutions designed for rapid scaling have ignored fundamental security steps, especially in onboarding flows, document validation, and continuous monitoring. The result has been fertile ground for sophisticated attacks.
When fraud occurs, it affects the employee who loses a benefit, the small business that has its payments blocked, the manager who needs to justify losses, and the team that deals with rework, not to mention the financial losses—all impacted by a system they should trust. Ultimately, fraud is about the security of those who depend on these resources daily, and it was precisely this human element that made 2025 so challenging, amplifying the feeling of vulnerability among users, managers, and operators.
What to expect from 2026 and why there are real reasons for some optimism.
Despite the challenges, 2026 is likely to be a turning point, because the sector has finally understood that there is no alternative but to profoundly change the prevention architecture, as well as alliances between anti-fraud areas (even those of competing players). The movement that is beginning to consolidate involves three pillars:
– Predictive, not reactive, prevention: The sector is migrating to continuous behavioral analysis models and real-time (or transaction-time) anomaly detection. The trend is for risk decisions to occur in milliseconds, and this requires a revised, robust, and well-defined architecture. This shift in approach is the central point of maturity that the sector must achieve by 2026.
– Unified data orchestration: Fragmentation will be replaced by platforms that connect identity, usage, transactions, geolocation, profiles, and relationship history. The unified view will allow the detection of patterns that previously went unnoticed. The key word is: interoperability.
– Secure experience without friction: The great achievement of 2026 will be balancing rigor and fluidity, making some verifications invisible. Continuous authentications, intelligently validated documentation, and monitoring without interfering with the user journey are likely to become the rule, not the exception.
Fraud will continue to evolve, unfortunately, but 2026 is likely to be a year of silent, yet profound, reconstruction for defense teams. This will prioritize architecture reviews, further refinement of suppliers and companies that can better fit into these more innovative flows, and above all, a constant pursuit of creating more human, simple, educational, and secure financial experiences.
*Luanna Vargas is a payments intelligence specialist with over 19 years of experience in analytics, data science, fraud prevention, and artificial intelligence. With a degree in Mathematics and Computer Science from USP and a master's degree in Economics from UFSCar, she has worked at companies such as Visa and Will Bank, and was selected as an entrepreneur by Antler. She is currently the co-founder of HeronPay.
Notice: The opinion presented in this article is the responsibility of its author and not of ABES - Brazilian Association of Software Companies
