Financial firms tell Congress why they need AI

Img

Witnesses from companies across financial services and housing, including a credit union and a core services provider, attested to the value of AI in fighting fraud and boosting customer service at a hearing in the House of Representatives led by Patrick McHenry.
Al Drago/Bloomberg via Getty Images

Representatives from companies spanning financial services and housing testified on Tuesday about the importance of high-quality data and auditability when applying artificial intelligence across their industries, as well as its promise in fighting fraud and elevating competition among firms. 

The testimony, part of a House Financial Services Committee hearing about the benefits and risks of AI, follows the release of a report on June 18 by the committee's task force covering how regulators and financial firms use AI. 

"To be sure, AI is not new to the financial services industry," said Chairman Patrick McHenry, R-N.C., in his opening remarks on Tuesday. "Financial firms have used algorithms for decades, whether analyzing large datasets or trading. However, new developments in generative AI have excited the imaginations of many."

For instance, Nasdaq is deploying AI to boost the effectiveness of financial crime risk management at financial institutions and machine learning to detect insider trading, market manipulation and fraud.

"The adoption of AI technology does not necessarily demand sweeping regulatory changes," said John Zecca, global chief legal, risk and regulatory officer at Nasdaq, during the hearing. "New regulation should avoid focusing on specific technology itself and instead be proportional to the benefits, risks and harm of the AI application. The regulation of an AI tool to detect drug trafficking should not be the same as the regulation of an identical AI tool if it is used to approve apartment rental applications."

One focus was on how AI lets community financial institutions, including the $1.6 billion-asset Great Lakes Credit Union, improve service to its members on a level that can compete with larger firms. 

Great Lakes Credit Union deployed a voice-enabled virtual assistant, Olive, with vendor Interface.ai in 2023. Olive handles 60% of inbound calls during business hours and 75% after business hours, reported chief operating officer Elizabeth Osborne during the hearing.

As a result, the Bannockburn, Illinois-based credit union's call center agents have transitioned from transactional support roles to more consultative and advisory roles, with higher pay grades and better career paths. The next step is to upgrade Olive to communicate in Spanish, a language spoken by a large percentage of Great Lakes' membership base. 

Reps. Patrick McHenry, R-N.C., and Maxine Waters, D-Calif., who run the House Financial Services Committee, today released their report on artificial intelligence in the financial industry.

July 18

"AI helps us to enhance service to our members that economies of scale might not otherwise allow us to do," she said.

Great Lakes is focusing on customer service use cases for now, and does not apply AI to lending decisions or to detecting and preventing fraud. But it is exploring use cases in both of those areas.

Another question was how to ensure compliance with regulatory and ethical standards, including for firms that are heavily reliant on third parties such as Great Lakes.

"Compliance is at the forefront of our customers' minds," said Frederick Reynolds, deputy general counsel for regulatory legal and chief compliance officer at core technology provider FIS. 

He said during the hearing that FIS quickly set up a governance council as part of its AI practices. 

"The key is auditability, making sure we are clear as to how and why we are using AI and what controls we have in place," he said. 

FIS is conscious that it is the "first line of defense" when deploying technology to its customers, which are predominantly banks and brokerages, said Reynolds. "We provide information so they can make their assessments to ensure it is operating as designed and without bias."

Keeping a human in the loop is "critical" for the technology at this stage, he said, to ensure AI systems "learn" from their mistakes and course-correct for fairness.

On its end, FIS is focusing on the impact of technologies including generative AI on fraud prevention, identity verification and existing customer due diligence procedures. 

"As these technologies become more widely adopted, our traditional means of assessing trust, that is identity, are less effective," said Reynolds. "It's imperative for the next generation of customer due diligence and identity controls to take advantage of generative AI and Web3 while also protecting against misuse."

Osborne said her credit union monitors its third parties closely and goes through a rigorous process to vet its partners, including their reputations, histories of data breaches and financials. 

"We treat them as a critical vendor," she said. 


More From Life Style