Augmented Analytics Helps Banks Weed Out Fraudsters and Reduce False Positives

The fraudsters get sophisticated, and the schemes get faster and evolve with a fluidity that’s breathtaking. 

Need proof? In the last year, $50 billion was lost to identity theft and bank fraud.

The banks, then, have to step things up a bit — to augment the lines of defense they’ve erected against the criminals.

As banks’ clients — both commercial ones and consumers too — make the shift from ACH and wires, from cards to digital payments, a holistic approach to viewing transactions, and anomalies is critical.

See also: Digital Fraud Tracker: Explaining Third- and First-Party Fraud

An incremental approach is necessary, a trio of executives from the financial services sector told PYMNTS — including augmented analytics.

The panel included Featurespace Head of Americas/U.S. AML and Financial Crimes SME Todd Raque, Texas Capital Bank SVP of Enterprise Fraud Operations Jamie Burud and Webster Bank SVP of Fraud Services Kevin Thompson.

As Raque said, the rise in fraud, in attacks on banks and financial institutions (FIs) in general has been unprecedented — evidenced by 1,300% jumps in ransomware attacks — and is set to increase.

In the great digital shift, he said, “some organizations were well-prepared, and some have been trying to play catch up.”

Augmented analytics solutions allow organizations to automate their fraud detection and management operations, utilizing artificial intelligence (AI). Those advanced technologies help weed out potential issues, eliminating the need to do so manually.

But this is no rip-and-replace strategy. That augmented approach, said Raque, takes FIs’ existing capabilities and overlays them with new capabilities, from a technology and tool perspective.

“You take what’s working today and bring additional focus and capabilities to your fraud-fighting teams,” he said.

“Some banks are more mature or in a better position to leverage machine learning capability,” said Raque. But no matter the size of the institution or the type of organization (regulated or unregulated in the payments and financial services arena), all firms are trying to figure out how machine learning will fit into their program.

Simply put, he said that machine learning brings capabilities to the battle against bad actors that legacy systems cannot provide.

“Machine learning,” he said, “can look across larger, more holistic data sets in ways that add context.”

And context, he said, is what is missing from the rules-based systems — with the added benefit of lowering false positives.

Burud maintained that the augmented approach is “absolutely spot on, especially as we see the fraud landscape continue to change.” Anti-fraud solutions, she said, need to be scalable as banks seek to protect their clients.

Webster Bank’s Thompson said augmented analytics has helped give rise to higher quality fraud alerts. Bringing in more data across various channels helps banks gain a holistic view of customers and clients.

Better Client Experience  

“This approach has helped us identify fraud, and create a better client experience as well.” At a high level, he said, augmented analytics allows banks to leave clients “alone” when there is no need to talk with them, but gives them the insight they need to reach out and engage with those customers if there is anything that looks suspicious.

Featurespace’s Raque stated that a number of advanced technologies are coming to the forefront in the battle against fraud, chief among them machine learning. But the uptake may be a bit uneven, at least for now.

As of now, said the panelists, there is no silver bullet, no single solution that battles every type of fraud that is out there (or that is coming down the line).

Said Webster Bank’s Thompson, “you’ve got to have a layered approach” that might take bring together the products and services offered by a number of vendors.

No matter which vendors are chosen, FIs need to assess risk management frameworks and processes, with both fraud and AML in the crosshairs (fraud and AML, after all, are becoming ever more intertwined).

See also: Why Context Is Key to Effectively Fighting Fraud

Burud said that assessing vulnerabilities needs a collective effort between AML teams, fraud teams and cybersecurity teams.

“Maybe you don’t see so much of one type of fraud today, but the possibility is there. So what is your risk appetite for managing that across the board? That assessment needs to be done across teams internally,” he said.

Looking ahead, Thompson said, the biggest threat that lies ahead rests with scams — targeting consumers, perhaps, or business email compromise efforts that have commercial targets.

“It’s getting harder for the bad guys to ‘get into’ the banks but it’s getting easier to trick people into giving money away, he noted. FIs need to forge ahead with customer outreach and education, especially as money moves faster through Zelle and other conduits, said the panelists.

Education is foundational, said Raque, who added that “we need to work within institutions, need to see work between institutions, and private-public partnerships to share what we’ve learned.”