red modern music youtube channel art (5)

Breaking down silos: Fraud detection using consortium data and big data analytics

Have you ever noticed how much harder it is to achieve your goals at work when each team is working in a silo? In the absence of collaboration and information sharing it’s almost impossible to have the same breadth and depth of impact as you can when your teams work together.

The same idea holds when it comes to gleaning insights from shared fraud data across financial institutions.

As fraud has proliferated, especially via online channels in the post-pandemic digital world, it has become more important than ever for banks and credit unions to work together to detect and stop fraud leveraging consortium data and analytics. 

Fraud prevention starts with identity verification

Challenges with identity verification are often the catalyst for today’s most pervasive fraud issues. Losses from identity fraud reached a staggering $43 billion in 2022 according to Javelin Strategy & Research.1 

Once a criminal opens a bank account with stolen or fabricated credentials (e.g. a synthetic identity), that opens the door for them to commit other financial crimes like money laundering or credit card fraud.  

Stopping fraudsters before they enter the system is key to preventing a wide range of follow-on crimes – and fraud detection within networks using big data analytics can help banks and credit unions do just that.

By looking at key identity attributes, behavioral patterns and anomalies, machine learning models derived from consortium data can predict the likelihood that someone truly is who they claim to be in real-time.

And even if fraudsters get past the first line of defense, predictive modeling can help financial institutions prevent additional fraudulent transactions. 

Take check fraud for example, a $24 billion problem for banks and credit unions in 2023.2 

A fraudster attempting to deposit a counterfeit check may already have an open account in good standing with the bank, or a good customer may unwittingly attempt to deposit a bad check after falling victim to fraud themselves. 

In either case, the customer probably wasn’t flagged during the identity verification process – but with predictive modeling on top of consortium data, their financial institutions can still be alerted to the fraudulent check and prevent it from being deposited at the point of presentment.  

Predictive modeling powered by consortium data plays a critical role in preventing fraud across the financial ecosystem

As Chief Data Officer at Early Warning®, I am constantly monitoring the evolving nature of financial fraud and safeguarding the crucial role that consortium data plays in protecting the integrity of our financial systems.

You can think of consortium data as pieces of a puzzle. Each bank or credit union holds a unique puzzle piece representing insights into fraud and transaction activity occurring at their institution. When these pieces are shared and combined, a comprehensive picture of fraud activity across the financial landscape emerges. It’s akin to assembling a collective defense strategy where the whole is more powerful than the sum of its parts, providing a robust shield against financial fraud.

Consortia are incredibly powerful for fighting financial crime because when an individual commits fraud at one institution, it’s made known to all participating institutions across the financial system. The historical data about these occurrences of fraud are then used to inform machine learning models that help predict the likelihood of fraud in future situations. 

This collective intelligence is instrumental in identifying patterns, anomalies, and trends that might go unnoticed in the confines of individual datasets.  

So how can we use big data analytics for fraud detection and prevention? 

Big data analytics involves the analysis of vast datasets to extract valuable insights. In the context of fraud detection, this means scrutinizing enormous volumes of personally identifiable information, transaction data, and user behavior patterns over time. 

The ability to process and analyze data at such scale and speed empowers consortia and their various financial institution members to identify unusual or suspicious activity and potential fraud with greater accuracy. Some of the key benefits include: 

  • Pattern recognition and anomaly detection: big data analytics can be used to identify patterns of normal behavior and flag any activities that deviate from established patterns. 
  • Real-time monitoring: leveraging big data analytics for real-time monitoring of transactions enables financial institutions to respond immediately to unusual patterns, reducing the impact of fraudulent activities.  
  • Machine learning algorithms: machine learning models can be trained on historical data to predict and prevent future fraud and are continuously updated to adapt to changing fraud tactics. 

One of the primary advantages of consortium data lies in its breadth and depth

A panoramic perspective is crucial in an environment where fraudsters are constantly coming up with new strategies and tactics to exploit any gap in our defenses. 

In the case of Early Warning, our consortium, the National Shared DatabaseSM resource, collects account-holder data contributed by 45 out of 50 of the largest U.S.-based financial institutions, providing broad consumer coverage and high reliability.  It’s recognized as a specialty consumer reporting agency and is regulated by the Office of the Comptroller of Currency and the Consumer Financial Protection Bureau (CFPB).

We employ a “give-to-get” model in which financial institutions that consume our products must also contribute account-holder data such as demand deposit account (DDA) data and personally identifying information (PII), to the consortium. This in turn increases the consortium’s coverage and ultimately strengthens the data models it powers.  

Our predictive models are trained on data from 697 million deposit accounts, and this comprehensive coverage means we can provide meaningful responses on about 94% of deposit inquiries. By applying sophisticated machine learning models to our vast pool of customer and transaction data, we’re able to more accurately identify and differentiate genuine transactions from fraudulent ones. 

Early Warning® has been providing shared insights and fraud detection solutions to financial institutions for over 30 years. It originally started as a way for banks and credit unions to share information on deposit fraud and has now evolved into a consortium with comprehensive data on identity, payment, and deposit fraud from over 2,500 participating banks and credit unions. 

The fight continues, but the future is bright

Ultimately, the combination of big data analytics and consortium data represents a formidable alliance in the ongoing battle against fraud. As technology continues to progress, so too will the strategies employed by financial institutions to protect and maintain the trust of their customers. Through collaborative efforts and the harnessing of advanced predictive modeling tools, banks are poised to stay ahead of the ever-changing landscape of financial crime and better protect consumers from becoming victims of fraud and scams. 


  1. Javelin Research & Strategy, 2023 Identity Fraud Study: The Butterfly Effect, March 2023
  2. American Bankers Association, Back with a vengeance: The challenges of check fraud, March 2023
Tagged with:
Posted in: , ,
Author: Francesca La O

Francesca La O is the Chief Data Officer at Early Warning®, a financial services technology leader that has been empowering and protecting consumers, small businesses, and the U.S. financial system with cutting-edge fraud and payment solutions for more than three decades. As CDO, Francesca leads enterprise-wide data and analytics strategy and execution. Her teams mine insights, build internal and external data products, and define robust programs to manage data that drives enterprise outcomes. Her team's work empowers 2,500 US financial institutions to verify accounts and stop fraud. Francesca has over 20 years of experience in data, analytics, e-commerce, and strategy, with a track record of delivering impactful business outcomes within the technology space.