Fraudsters Share Playbooks, So Why Won’t Banks Share Data?

Watch more: Entersekt Warns Banks Can’t Beat Crime With Old Data Alone

In the battle against financial crime, data is both the sword and the shield. Fraudsters are faster, more sophisticated and more global than ever. Banks, payment providers and FinTech firms must respond with a mix of vigilance, agility and intelligence, built on information that is not just abundant, but actionable.

Pradheep Sampath, chief product officer at Entersekt, argues that data — traditional, alternative and shared — has never been more important in preserving trust in financial transactions.

But he also warns that the fight requires a multipronged approach that blends legacy data with newer, faster signals.

“It’s a team sport,” he said as part of the PYMNTS “What’s Next in Payments” series on reliable data in the service of financial services firms. “And the thread that binds us all together is data that’s actionable, shared in good faith, and governed responsibly.”

Historically, financial institutions have leaned heavily on data from government and regulatory sources in the U.S., such as the Financial Crimes Enforcement Network’s suspicious activity reports, the Federal Trade Commission’s Consumer Sentinel alerts, the Office of the Comptroller of the Currency’s fraud bulletins, and the Federal Reserve’s payment fraud reports.

Cross-border, the industry has tapped alerts from agencies like Homeland Security, Interpol and Europol.

Read more: Entersekt CEO on Scams: ‘Bigger the Bullet, the Thicker the Armor’

“These data feeds,” Sampath said, “have become ingrained in the sector’s sources and methods for assessing transaction efficacy and customer onboarding.”

But reliance on these sources alone is risky, particularly as their timeliness, completeness or accessibility may come under pressure. The challenge is to supplement these official channels with richer, faster industry data — from card network alerts to ecosystem-wide behavioral trends.

“We need to detect suspicious trends faster by looking beyond just government data feeds,” he said.

 

 

Blending Traditional, Alternative Data

Sampath said traditional, historical data remains a cornerstone of fraud prevention. “You’ve got to look back to look ahead,” he said. “Historical data is the foundation — any model you have today needs that history. But we also need real-time risk radars because threats evolve every week.”

These newer “radars” include transaction-driven insights, behavioral signals, device fingerprints, and geolocation patterns.

The goal is to blend these disparate signals into a more responsive, adaptive fraud defense.

“Looking back can’t always give you answers to evolving threat vectors,” he said. “You need both accuracy and speed — protecting legitimate users while quickly identifying emerging fraud patterns.”

Artificial intelligence (AI) has become a central tool in this effort, but Sampath cautions against uncritical adoption. “It’s a strong yes to AI, but with guardrails,” he says. “AI models must be explainable, not a black box. We need governance, bias checks, and continuous monitoring to make sure they perform the way they’re supposed to.”

Guardrails, Governance and Consortium Data

Guardrails start with understanding the provenance of data. “You’ve got to have some idea of the chain of custody,” Sampath told PYMNTS.

This includes robust contractual agreements with data providers, adherence to regulations, and practices like data minimization and anonymization. Tools such as metadata management and data lineage tracking can verify sources and ensure compliance.

Sampath also sees promise in the consortium model — shared fraud data ecosystems where competitors cooperate for mutual protection. “Where the industry needs to go forward,” he said, “is to share data in a responsible manner across consortia, while preserving fair competition.”

Achieving that balance requires privacy-enhancing methods such as encryption and federated data models, along with common governance and trust standards. This approach, he argued, will enable legitimate users to enjoy smooth transactions while keeping ahead of emerging fraud vectors.

Minimalist Data Sharing

One of Entersekt’s key focus areas is minimalist data sharing, transmitting only what is strictly necessary to validate a claim or detect fraud. Sampath offers a simple example: verifying that a customer is over 21 to purchase alcohol. “They don’t need your full name, date of birth, address, or whether you’re an organ donor,” he said. “The fact that you are over 21 is the only claim that needs to be asserted.”

This principle extends to consortium fraud data. “It’s possible for us to share that this IP, this device, this transaction pattern, or these geolocations are risky — without revealing the actual user, transaction, vendor or card network,” he said. “The whole community benefits from these signals without compromising privacy or exposing personal identifiers.”

For Sampath, this targeted approach represents the future of fraud prevention — one that respects privacy, strengthens trust and improves speed in catching bad actors. “The decisions based on this data must be sound, fair, and able to withstand the pressure of emerging threats,” he says. “That’s how you keep fraud at bay.”

Fighting fraud, Sampath says, is no longer the job of individual institutions. “It takes a village to fight fraud,” he said, “and it’s heartening to see competitors come together to fight the good fight.”

The post Fraudsters Share Playbooks, So Why Won’t Banks Share Data? appeared first on PYMNTS.com.