2025 Fraud Prevention Trends: End-of-Year Scorecard

Last year, we peered into our crystal ball to predict the major trends in fraud and financial crime that would play out in 2025. As the year wraps up, it’s time to assess if our initial expectations for 2025 came true or if we missed the mark.

Read on and look back at Feedzai’s initial 2025 forecasts and how accurately these predictions panned out. 

Key Takeaways

• “Failure to Prevent” fraud initiatives are taking effect as financial institutions move from leaving victims to bear the burden of scam losses to having financial institutions and other players assume greater responsibility. 
• Intelligence-sharing efforts proved an important factor in 2025, including Feedzai’s collaboration with Mastercard.
• Financial institutions are making progress in catching high-risk transactions, but the speed of change has been slower than expected.
• Digital currency initiatives have moved from theoretical concepts to becoming legitimized by legislation and official policy moves.

5 Fraud and Financial Crime Predictions in 2025: The End of Year Results

Which predictions came true and which ones were off? Let’s explore.

The Rise of ‘Failure to Prevent’ Fraud Regulations

Original Prediction (Jan. 2025): In 2025, we’ll see “failure to prevent” fraud initiatives and victim-focused reimbursement rules placing heavier burdens on financial institutions. The UK is a prime mover here, setting standards that will likely echo around the globe. Financial leaders, facing heightened scrutiny, can’t just pay lip service to security; they’ll need to show their work. This translates to more robust technologies, well-trained teams, and stronger oversight across the board. The ripple effect goes beyond the boardroom. This raised bar means cutting off illicit funding streams to criminal enterprises, striking at the heart of human trafficking and drug-related crimes. Institutions aren’t just protecting their reputations—they’re helping restore confidence in financial systems and, in doing so, making the world a bit safer.

Result: Came True

As predicted, we are seeing a shift away from shouldering victims with scam losses toward financial institutions assuming greater responsibility. We expected regulations from the UK’s Payment Services Regulator (PSR) requiring liability scam losses to be split evenly between sending and receiving banks to set the pace for other regions. Several major regions passed their own scam liability frameworks, raising the bar for protecting customers and mandating proactive measures to build trust in the financial system. 

Here’s how other regions have followed suit. 

• EU: Under PSD3 regulations, financial institutions are expected to assume liability for certain types of scams. This is especially true of bank impersonation scams in which bad actors pretend to represent a customer’s bank to defraud them.

• Australia: Australia’s Scam Prevention Framework (SPF) is making cross-sector collaboration in fraud prevention a reality by bringing together banks, telecoms, digital platform providers, and more to take “reasonable steps” to detect, prevent, report, and respond to scams, or face civil penalties. This is a significant measure that brings in equally essential players, not just financial institutions, safeguard for consumer safety.

• Singapore: The Monetary Authority of Singapore (MAS) recently implemented the Shared Responsibility Framework (SRF), which expands liability for scam losses from banks to telecom companies. 

Collaboration and Intelligence Sharing in Fraud Prevention

Original Prediction (Jan. 2025): In the past, many efforts to combat fraud have been scattered and siloed. But 2025 is the year true collaboration takes off. Projects like the FIRE pilot—where banks and tech companies share intelligence—show the power of uniting under one mission. We will also see the network effect bringing value as payment providers around the world recognize their unique perspective of the data they hold.

By aligning strategies and data, organizations can prevent suspicious activity upstream before it turns into transactional fraud. This harmony pays dividends in real-world outcomes: customers see more consistent safeguards across services, and the industry as a whole can elevate standards while staying nimble. As cooperation replaces competition in the fight against crime, we open doors to smarter detection and a healthier financial ecosystem for everyone.

Result: Came True

As predicted, intelligence sharing was a notable factor in the past year, including Feedzai’s groundbreaking partnership with Mastercard. Under this collaboration, Feedzai is helping Mastercard to leverage their position as a payment network to derive insights and help member banks more effectively prevent fraud.

In addition to greater network intelligence and collaboration, we’re also seeing increased engagements between two historically different disciplines: cybersecurity and fraud prevention as “Cyber-Fraud Fusion.” Fraud teams, for example, have extensive context on authorized payments and customer behavior but lack insights into activity before authentication. Meanwhile, cybersecurity teams can view network-wide traffic, including activity beyond a bank’s domain, but have blind spots regarding customer behavior or downstream payments.

Under the “Cyber-Fraud Fusion” concept, members of these different teams are considering merging operations and data insights to build a more unified defense against fraud. 

AI in Fraud Detection: Good vs. Bad

Original Prediction (Jan. 2025): If 2024 was the year we learned how AI can supercharge both our defenses and the fraudsters’ offenses, 2025 is when we put that knowledge into practice. Artificial intelligence might create deepfakes and voice clones that fool the unsuspecting, and it can churn out synthetic identities at scale. However, it can also become a powerful ally: helping investigators identify patterns in oceans of data, shining light on anomalies before they become crises, and speeding up the time-to-value of analytical responses. 

In 2025, we’ll see FIs move GenAI from proofs of concept into full-scale production environments. These institutions will rely on advanced AI capabilities to optimize operations, bolster security controls, and respond to fraud threats with unprecedented agility. The upcoming year promises not just incremental improvements but a transformative leap toward a more intelligent and proactive approach to fraud prevention. 

Result: Not Exactly True

This prediction didn’t quite pan out as we expected. In some ways, we were right in the spirit of our prediction, but we were off on the timeline as to how events would unfold.

For example, we didn’t see widespread adoption of agentic or generative AI across fraud teams within financial institutions. In reality, while banks are using Generative AI, most activity is focused on providing internal-knowledge-based answers and acting as a co-pilot for day-to-day tasks.  We didn’t see many direct, consumer-facing fraud prevention use cases applied, and adoption in fraud operations was also limited.

While Agentic AI is being explored actively in financial services, many early proof of concepts (POCs) fell short, according to research from the Massachusetts Institute of Technology.¹ The key to success will be connecting agent pilots directly into existing workflows; this will drive adoption and deliver the most value, with the automation of data summaries within case management solutions seeming to be the most common bet among banks. Overall, the technology is advancing, but a larger-scale production and adoption of autonomous AI agents will take longer than we initially thought.

Scam Prevention in 2025: Breaking the Spell

Original Prediction (Jan. 2025): One of the toughest battles is against scams where the victim, under psychological pressure, pushes through fraudulent transactions—even as security teams try to intervene. In 2025, we’ll see financial institutions invest in more nuanced training for their staff and deploy advanced tools, including AI-driven assistance, to break these “spells.” Banks can lighten these crimes’ emotional toll on individuals and communities by helping frontline teams navigate complex human behaviors and empowering customers to recognize warning signs. The result is a more resilient defense system that doesn’t just rely on stopping bad actors but also on fostering confidence and clear thinking in the very people those criminals hope to ensnare.

Result: Not Exactly True

Progress has been made in areas like aligning processes and deploying better tooling. But the speed of change was slower than we expected, and innovation has not been rife in this area; the problem is far from solved. 

The good news is that financial institutions are getting better at catching high-risk transactions and intervening when customers appear to be under a scammer’s influence. 

However, there’s still work to be done. Trying to break months of trust that has been established between victim and fraudster during a short phone call with the bank is a complex challenge; it will require substantial investment in training financial institution staff and investing in alternative technology and data solutions.

Digital Currencies: Riding the Next Wave of Demand

Original Prediction (Jan. 2025):  The resurgence of interest in digital currencies—sparked by Bitcoin’s price recovery and the exploration of projects like the Digital Euro—will push banks and Neo Banks to reassess their risk strategies. For traditional financial institutions, it’s a moment to consider whether a blanket ban on crypto-bound transactions still makes sense. For newer, more agile Neo Banks already weaving these offerings into their core services, 2025 will mean tightening controls and reevaluating how best to mitigate money mule activities. It’s a delicate balancing act: meeting consumer demand, exploring innovative technologies, and doing so responsibly. Done right, managing digital currencies can lead to stronger fraud defenses and more confident, forward-looking financial brands.

Result: Came True

As originally predicted, we have seen digital currency move from theoretical discussions to becoming legitimized through legislation and official policy. For example, when Bitcoin saw a significant price surge earlier this year,² it encouraged interest from casual investors (before the price dropped considerably again).³ 

Beyond Bitcoin, digital currency has seen several major steps toward becoming legitimate in the past year. First, in the US, the “GENIUS Act” was enacted, becoming the nation’s first major initiative to regulate stablecoins.⁴ The Act limits issuance to banks or non-bank entities, requiring all stablecoins to be backed by 1:1 reserves of cash and low-risk assets, and ensuring compliance with the Bank Secrecy Act to implement strong anti-money laundering and consumer protection measures.

In Europe, work on the Digital Euro project continues. The European Central Bank has completed its preparatory phase and is seeking to be ready for issuance by 2029. This initiative could see the digital euro become legal tender linked to the euro, boosting financial inclusion for the 440 million citizens in the eurozone.

All in all, while we didn’t have a perfect record for our 2025 fraud and financial crime predictions, we were very close. As we prepare for a new year, we’ll be keeping a close eye on new trends that will shape the world of financial services. More importantly, we’ll show you the steps you can take to prepare to respond to new developments when they unfold. 

Additional Resources

• Blog: What Is Digital Currency? Types, Risks, and How to Protect Yourself
• Webinar: The Future of Fraud is Here: Are You Ready?
• Analyst: How Data Orchestration Unlocks Safer, Faster Account Opening 
• Solution: AI that Goes Beyond the Ordinary

This article was first published by American Banker.