The International Monetary Fund Warning: Fraudsters Thrive on FIs Not Sharing Data
- I fraud thrives when fragmented data keeps financial institutions in silos.
- IMF urges banks to share transactional data so AI can spot patterns..
- Open data consortiums and check image sharing strengthen deposit fraud defenses.
PYMNTS.COM reports on The International Monetary Fund’s recent warning to banks, which is simple: AI can help fight fraud, but only if institutions are willing to share data. That message applies just as strongly to check fraud, where banks still operate with limited visibility into what happened before a deposited item reaches their system.
Fraud has always thrived on information gaps. Today, criminals can move faster because AI helps them scale impersonation, document abuse, and coordination across multiple institutions, while banks often keep the data that could expose those patterns locked inside separate systems. Check fraud has the same structural weakness: one bank may see only a single deposit, not the broader pattern of the same bad actor, altered item, or stolen check showing up elsewhere.
IMF argues that the industry doesn't need better tools, it needs better visibility.
Why the IMF Warning Matters
The IMF argues that AI and machine learning work best when they have access to large, diverse datasets, and that fragmented, siloed environments limit what those models can detect. It also points to interoperability, APIs, and common standards as the path to better collaboration across institutions. That is a useful framework for understanding why fraud prevention is moving away from isolated defense and toward shared intelligence.
For banks, the challenge is not just technical, it's also cultural and competitive because institutions are used to protecting data rather than pooling it. But fraud is now cross-institutional, and the people committing it are already behaving that way.
The Check Fraud Parallel
Check fraud has a very similar problem: Banks and processors often do not share enough data, and they usually do not share check images broadly enough to build a fuller fraud picture. Without that broader view, one institution may approve or process an item that another has already seen as suspicious, counterfeit, or altered. That makes check fraud especially hard to stop when the fraudster is moving from one institution to another.
Many FIs are turning to leveraging consortium data to address deposit fraud. However, one factor being overlooked is whether the consortium is opened or closed.
An open consortium allows more than just financial institutions to contribute data. Other check processors such as service bureaus can both contribute data and leverage the consortium’s intelligence, which broadens coverage and increases the chance of spotting suspicious patterns. An open model is designed to create wider visibility across the payment.
A closed consortium is limited to financial institutions that have joined the network. That can make governance simpler, but it also narrows the amount of data available and can leave blind spots if important processors or intermediaries are excluded. In fraud prevention, fewer participants usually means fewer signals, and fewer signals mean weaker detection.
Additionally, consortium data is most powerful when it includes not just transactional data and account historical data, but also images good items. Check fraud is not only about account history; it is also about the physical and visual evidence in the item itself. If the industry does not share this data, it loses one of the best ways to catch fraud early.
Why Data Sharing Matters in Check Fraud
That's why data-sharing consortiums matter. Collaboration can reveal patterns no single institution can detect on its own, especially when the consortium includes both transactional data and check image intelligence.
Criminals are already sharing tactics, reusing instruments, and exploiting gaps across institutions, while banks have historically treated data as a competitive asset instead of a collective defense layer. This is starting to change as more institutions recognize that fraud is an ecosystem problem, not a single-bank problem.
Data consortiums are an important component -- or layer -- for any financial institution's broader check fraud strategy. By combining this data, along with analyzing the images of checks, banks can achieve over 95% detection rates and reduce fraud losses.