Grocery Stores Protest NYC’s Proposed Restrictions on Facial-Recognition Software

The tech is needed to curtail shoplifters and keep stores open, business owners say

Facial recognition used on pedestrians on a New York Street.Credit: Getty Images

Abill to restrict the use of facial recognition technology in New York City is facing opposition from grocery store owners, who say that the tool is helping to curtail rising shoplifting.

The proposal would, in part, require businesses to inform customers and obtain written consent before using biometric recognition technology under penalty of a $5,000 fine for violations.

Opponents of the New York City Council bill testified at a hearing last month that the technology is helping small businesses fight back against theft.

“In the past, we would take Polaroid photos of thieves and Scotch-tape them by the time clocks, so employees could recognize the thieves if they enter the supermarket again,” testified Avi Kaner, co-owner of the Morton Williams supermarket chain. “There is no difference between that and using facial recognition.”

“If a thief is caught on camera wiping out an entire section of the supermarket, we must have the ability to prevent that thief from coming into our supermarket again.”

The chain has seen its profit margins cut by 2% within the past year due to theft, and spent over $1 million to hire off-duty NYPD officers to work security, Kaner said.

“This has practically wiped out any profit we have and threatens to shut down a number of our stores,” he testified.

“We are not collecting biometric data,” Kaner said. “We are simply using photos of known thieves to prevent their entry into our stores.”

Jay Peltz, the general counsel and senior vice president of government relations for the Food Industry Alliance of New York State trade association, agreed, saying that without tools to curtail theft, stores could be forced to close, widening food deserts.

“The inability to collect biometric identifier information and use biometric recognition technology would seriously undermine the ability of the city’s grocers to deter shoplifting and assist law enforcement investigations of repeat offenders,” Peltz testified. “The failure to reverse rising thefts at marginal grocery stores will likely result in the closure of those locations, thus exacerbating the city’s food desert problem.”

Facial recognition technology also helps reduce potentially violent confrontations between store workers and would-be shoplifters, Peltz contended.

“A rise in retail theft is accompanied by an increase in threats of violence and actual violence during the commission of such crimes,” Peltz said. “This creates the need for merchants to use legal, ethical methods that are not confrontational to deter theft and assist law enforcement investigations of repeat offenders.”

Proponents of the bill, however, say that the tech is invasive and has a history of struggling to differentiate people with dark skin as reliably as light-skinned people.

“Facial recognition and other biometric surveillance tools enable and amplify the invasive tracking of who we are, where we go, and who we meet,”

testified Daniel Schwarz on behalf of the New York Civil Liberties Union.

“The widespread use of these technologies presents a clear danger to all New Yorkers' civil liberties and threatens to erode our fundamental rights to privacy, protest, and equal treatment under the law.”

Jumaane Williams, the city’s public advocate, pointed to historic disparities in the tech’s accuracy making distinctions across complexions.

“There must be a strong deterrent to utilizing biometric technologies because the technology has proven to be inaccurate towards darker-skinned individuals,” said Williams.

He cited a 2019 finding by the Massachusetts Institute of Technology that

“facial recognition software marketed by Amazon misidentified darker-skinned women 31% of the time.”

Derk Boss, president of the International Society of Certified Surveillance Professionals, however, said in testimony opposing the bill that the tech’s accuracy has improved — and, arguably, less discriminatory than human-based security.

“FRT is machine-driven, it is not subjective like security guard wanding or other measures where racial bias may creep in,” testified Boss, adding that the National Institute of Standards and Technology has “​​found that FRT accuracy had improved dramatically and that more accurate systems were less likely to make errors based on race or gender.”

As the bill continues to make its way through the legislative process, it now has the sponsorship of 15 members of the 51-member Council.

The proposal was spurred by the use of facial recognition technology by New York Knicks and Rangers owner James Dolan at Madison Square Garden and other venues he oversees.

Dolan has said that he uses the tech to bar lawyers whose firms have pending litigation involving his holdings.

Source: The Messenger