Banks often don’t see a fraud scheme until money has left the account, and then it’s too late. A stolen check, compromised card or banking credential may already have circulated among criminals, giving fraud teams little time to intervene.
As payments move faster and fraudsters gain tools that make it easier to exploit stolen information, transaction monitoring remains essential.
But the transaction itself can arrive relatively late in the sequence of events that produced the fraud.
“In the life cycle of a fraud or a scam, most of those fraudulent scenarios are happening outside of the banking system,” Colin Parsons, head of fraud product strategy at Nasdaq Verafin, told PYMNTS. “The challenge really is that it only becomes visible to an institution at the time a transaction’s occurring, or money’s moving.”
Nasdaq Verafin and Q6 Cyber are trying to close part of that information gap through a partnership that brings intelligence gathered from online criminal activity into fraud workflows used by financial institutions. The initial focus includes compromised checks, payment cards and online banking credentials.
The premise is straightforward, although executing on it is not: Give an institution information about a compromised account before that compromise produces a transaction. Criminals are also adopting artificial intelligence, particularly in areas where stolen financial information can be turned into more convincing attacks.
“What you might find most surprising nowadays is the rate of AI adoption,” Q6 Cyber CEO Eli Dominitz noted during the same interview. Dominitz said Q6 Cyber is seeing fraudsters experiment with AI tools and large language models to automate existing practices, make social engineering faster and more tailored, and test ways to defeat financial institutions’ fraud controls.
That activity sits alongside the established trade in compromised checks, cards, bank accounts and other financial information.
A compromised check, for example, can contain the account holder’s name, account information and other identifying details. A stolen card or banking credential can similarly connect an individual or business to a particular financial institution. Dominitz said that information can help explain why subsequent impersonation attacks sometimes appear remarkably specific.
Financial institutions ask Q6 Cyber how a criminal knew that a particular person or business banked with them, he said. In some cases, the trail leads back to financial information that had already been exposed online.
The partnership joins two different views of the same fraud cycle. Q6 Cyber sees activity occurring outside the financial institution, while Nasdaq Verafin can bring those signals into an environment containing transaction and counterparty information from its financial institution network. Parsons said the combination gives investigators an external view alongside what they can see within their own institution.
Finding the Signal Before the Payment
Checks provide a measurable example of the opportunity created by earlier intelligence. The companies found an average of 10 days between Q6 Cyber identifying a stolen check offered for sale and the first fraudulent check being returned.
That interval gives a financial institution options it may not have after the check enters the payment system. Parsons said an institution could contact the customer or take stronger protective measures.
Parsons described the objective as moving institutions from dealing with the consequences of fraud toward preventing the transaction from occurring in the first place.
That doesn’t necessarily require more customer friction. Earlier intelligence can make intervention more selective because an institution has a specific reason to examine an account rather than applying another control broadly across transactions.
“You’re acting upstream, so you’re disrupting fewer transactions and allowing things to flow smoothly,” Parsons said.
Precision matters just as much inside the fraud department. Investigators already contend with alerts generated by transaction monitoring and other controls, making another stream of warnings useful only if it identifies activity worth investigating.
Dominitz said dark web intelligence can sometimes be tied to a particular institution and account rather than a general threat category. That specificity can change how an investigator prioritizes the warning.
“You also get to focus on the things that really matter and not spend too much time just dealing with noise,” he said.
Both sides are also applying AI to that workload. Parsons said Nasdaq Verafin uses machine-learning models with information from its consortium of more than 2,800 financial institutions to identify patterns associated with fraudulent scenarios. The company has also introduced agentic AI functionality intended to help investigators work alerts. Dominitz said Q6 Cyber is using AI to collect and process intelligence and produce alerts faster.
Yet technology gives fraudsters an advantage that financial institutions can’t comfortably copy: tolerance for error.
“We want be right 100% of the time,” Dominitz said. “They don’t have to be,” he said of the fraudsters, acknowledging that a small measure of success can give the bad actors a big payoff.
The post Nasdaq Verafin Takes the Fraud Fight to the Dark Web appeared first on PYMNTS.com.