U.S. banks deploy AI to monitor customers and workers amid tech backlash


Several U.S. banks have started rolling out camera software that can analyze customer preferences, monitor workers, and spot people sleeping near ATMs, although they remain wary of any backlash due to increased surveillance, more than a dozen banking and technology sources told Reuters.

Previously unreported trials at the City National Bank of Florida (BCI.SN) and JPMorgan Chase & Co (JPM.N) as well as previous deployments at banks such as Wells Fargo & Co (WFC.N) offer a rare view on the potential of the United States. financial institutions see facial recognition and associated artificial intelligence systems.

The widespread deployment of these visual AI tools in the highly regulated banking sector would be an important step towards their integration into American companies.

Bobby Dominguez, chief information security officer at City National, said smartphones that unlock through a face scan have led the way.

“We are already leveraging facial recognition on mobile,” he said. “Why not take advantage of it in the real world?”

City National will begin facial recognition trials early next year to identify ATM customers and branch office workers, with the aim of replacing awkward and less secure authentication measures at its 31 locations, Dominguez said. . Eventually, the software could spot people on government watch lists, he said.

JPMorgan said it was “doing a little video analytics technology test with a handful of branches in Ohio.” Wells Fargo said it worked to prevent fraud, but declined to discuss how.

Questions of civil liberties occupy an important place. Critics point to the arrests of innocent individuals over faulty facial matches, the disproportionate use of systems to monitor low-income and non-white communities, and the loss of privacy inherent in pervasive surveillance.

Portland, Ore., Since Jan. 1, has banned businesses from using facial recognition “in public accommodation,” and drugstore chain Rite Aid Corp (RAD.N) has shut down a national facial recognition program Last year.

Dominguez and other bank executives said their deployments were sensitive to the issues.

“We are never going to compromise the privacy of our customers,” Dominguez said. “We are starting very early to use technology that is already in use in other parts of the world and is quickly reaching the US banking system.”

Yet the big question among the banks, said Fredrik Nilsson, vice president of the Americas at Axis Communications, a leading manufacturer of surveillance cameras, is “what will the potential public reaction be if we deploy this?”

Walter Connors, chief information officer at Brannen Bank, said the Florida-based company discussed but did not embrace the technology for its 12 locations. “Anyone who walks into a branch expects to be registered,” Connors said. “But when you talk about facial recognition, it’s a bigger conversation.”


JPMorgan began evaluating the potential of computer vision in 2019 using software developed in-house to analyze archived images from Chase’s branches in New York and Ohio, where one of its two laboratories is located. innovation, said two people, including former employee Neil Bhandar, who oversaw some. of effort at the time.

Chase aims to collect data to better plan staff and design branches, three people said and the bank confirmed. Bhandar said some staff even went to one of Amazon.com Inc’s (AMZN.O) cashierless convenience stores to learn more about their computer vision system.

A preliminary analysis by Bhandar of the branch images found that more men would visit before or after lunch, while women tended to arrive in the middle of the afternoon. Bhandar said he also wanted to analyze whether women avoided compacted spaces in ATM lobbies because they might run into someone, but the pandemic disrupted the plan.

Testing facial recognition to identify customers when they walk into a Chase bank, if they have consented, was another possibility considered to improve their experience, said a current employee involved in innovation projects.

Chase wouldn’t be the first to evaluate these uses. A bank in the Northeast recently used computer vision to identify high-traffic areas at branches with newer layouts, an executive said, speaking on the condition that the company is not named.

A Midwestern credit union last year tested facial recognition for customer identification in four locations before stopping for cost reasons, a source said.

While Chase developed custom computer vision in-house using components from Google (GOOGL.O), IBM Watson (IBM.N), and Amazon Web Services, he also envisioned systems built entirely from startups. software from AnyVision and Vintra, said people including Bhandar. AnyVision declined to comment and Vintra did not respond to requests for comment.

Chase said he ultimately chose a different vendor, which he declined to name, out of 11 options being considered and began testing that company’s technology at a handful of Ohio locations last October. The effort is aimed at identifying transaction times, how many people leave due to long lines, and what activities workers occupy.

The bank added that facial, racial and sexual recognition were not part of this test.

According to some ethics experts, using technology to guess customer demographics can be problematic because it reinforces stereotypes. Some computer vision programs are also less accurate on people of color, and critics have warned that this could lead to unfair results.

Chase weighed the ethical issues. For example, some have called internally to reconsider planned testing in Harlem, a historically black neighborhood in New York City, because it could be seen as insensitive to racism, two people said. The discussions took place around the same time as a December 2019 New York Times article on racism at Chase Arizona branches.

Breed analysis was not part of the plans ultimately presented, and the Harlem branch was selected because it housed the other Chase Innovation Lab to assess new technology, people said and the bank confirmed.


The uses of security for computer vision have long attracted the interest of banks. Wells Fargo used the company’s primitive 3VR software over a decade ago to examine footage of crimes and see if any faces matched those of known offenders, said John Honovich, who worked at 3VR and founded the IPVM video surveillance research organization.

Identiv, which acquired 3VR in 2018, said bank sales were a major focus, but declined to comment on Wells Fargo.

A security official at a mid-sized southern bank, speaking on condition of anonymity to discuss covert measures, said that for the past 18 months he has deployed video analytics software in almost all branch offices to generate alerts when doors to safes, computer server rooms and other sensitive areas are left open.

Outside, the bank monitors strolls, such as the recurring problem of people setting up tents under the overhang for drive-through ATMs. Security personnel at a control center can play an audio recording politely asking those people to leave, the executive said.

The problem of people sleeping in closed ATM lobbies has long been an industry concern, said Brian Karas, vice president of sales at Airship Industries, which develops video management and analytics software.

Systems that detected loitering so staff could activate a siren or strobe light helped increase ATM usage and reduce vandalism for several banks, he said. While the companies did not want to move people seeking shelter, they felt it was necessary to make ATMs safe and accessible, Karas said.

Dominguez of City National said the bank’s branches use computer vision to detect suspicious activity outside.

The 2010 and 2011 sales reports reviewed by Reuters show that Bank of America Corp (BAC.N) purchased “iCVR” cameras, which were marketed at the time as helping organizations reduce loitering in counter lobbies. automatic. Bank of America said it no longer uses iCVR technology.

The Charlotte, NC-based bank’s interest in computer vision has not waned. Its managers met with AnyVision several times in 2019, most notably at a September conference where the startup demonstrated how it could identify the face of a Bank of America executive, according to tapes of the presentation seen by Reuters and one person present.

The bank said, “We are always looking at potential new technological solutions that are on the market.”

Our standards: Thomson Reuters Trust Principles.


About Author

Comments are closed.