In a meeting of the council’s public security committee on Friday, LBPD deputy chief Wally Hebeish defended his department’s limited use of facial recognition software, calling it a valuable tool only used to generate clues in criminal investigations. Even then, he said, it would need other evidence to back it up for detectives to take action.
“We don’t use this for mass surveillance of the community,” said Hebeish.
The three city councilors at Friday’s meeting – Suely Saro, Roberto Uranga and Suzie Price – took no direct stance on whether they thought the technology was appropriate, but some city officials have taken a firmer stance.
Since the beginning of last year, a two-person subcommittee of the city’s technology and innovation committee has been researching facial recognition technology published a report Long Beach recommends to ban its use for the time being and to create an opportunity to review new technologies before they are introduced.
“Based on previous research, the subcommittee finds that current face recognition technologies are not only inadequately accurate, but also pose significant and unequal risk for blacks and coloreds due to inherent algorithmic biases that have not been effectively addressed in software design,” says the Report. “The subcommittee is also concerned about the lack of independent review bodies to confirm that facial recognition technology is free from racial and other prejudice.”
It remains to be seen whether these recommendations will be accepted. The entire Technology and Innovation Committee will discuss the report next week.
How does the LBPD use face recognition?
Currently, the LBPD only uses facial recognition to generate clues in criminal investigations.
According to the department, the only method currently in use is to capture video from onlookers or surveillance cameras and compare them to a huge pool of approximately 9 million mug shots in the Los Angeles County Regional Identification System, or LACRIS.
LACRIS, administered by the Los Angeles County’s Sheriff’s Department, is a huge regional database that stores the identifying information of everyone arrested in the county.
The database, which the LBPD says has been used for years, states: its own limits what face recognition can be used for:
“All searches begin with a criminal investigation image (suspect), which is searched in a database of previously arrested people,” it says in their system manual.
If the system provides a match, it is not a fixed identification. Rather, “it supports the identification process by providing candidates as possible matches with the image sought,” according to LACRIS documents.
LACRIS searches were “instrumental” in identifying suspected looters after the protests against racial justice on May 31st Last year, LBPD’s Hebeish said at another meeting of the public safety committee last month.
But these leads from LACRIS, he said, are only a starting point.
“Our detectives are required to validate this information and give them directions to look for, and further investigation is needed to develop a reasonable suspicion and likely cause and prepare a case on record for prosecution,” Hebeish said.
Dozens of other law enforcement agencies in LA County also use the system. including the LAPD.
Nonetheless, civil rights organizations, including the American Civil Liberties Union, say the over-representation of black and Latin American men in wanted databases creates an inherently racist system.
In addition, the algorithms used in off-the-shelf facial recognition software are more widely used Identify people of color and women incorrectly—The specter of misguided investigations, more often targeting underrepresented communities.
The ACLU has joined a coalition of civil rights and data protection organizations urge a ban on government use of facial recognition technology.
In one Letter President Joe Biden said earlier this year that technology “is dangerous because it has clear racial, gender, and other prejudices, and it is dangerous when it works. Even if the technology worked perfectly, it would enable mass tracking of every person’s movements in public spaces – something unbearable in a free and open society. We cannot allow its normalization. “
On the spot, critics of facial recognition technology raised the same issues.
“Before the city gives the green light to the LBPD’s continued and unchecked use of facial recognition, the related issues of equity and racial justice must be addressed,” wrote Attorney Greg Buhl, who used public records to oppose the LBPD’s use of the technology.
Buhl claims that the LBPD at times marketed particularly controversial tools like those from Clearview AI – a company that scraped billions of publicly available photos on social media and other parts of the internet to build a huge database of identifiable information that it is now sent to sells subscribers.
Clearview AI’s product has sparked class actions in the US, government investigations in the UK and Australia, and an outright ban in Canada that it deemed illegal. according to the New York Times.
“It’s probably one of the most vicious companies ever created,” Buhl told the council’s public security committee last month.
At that meeting, Hebeish said his department had previously only used facial recognition software from vendors other than LACRIS on a trial basis, “but now we prohibit all attempts at this type of software without command-level authorization.”
At the Public Safety Committee hearing on Friday, half a dozen members of the public criticized the department’s use of facial recognition, pointing to the inherent biases highlighted by the ACLU, Buhl, and others.
But “research shows that residents generally support improved technology in police stations,” says the subcommittee’s report that will be debated next week in the technology and innovation committee.
The report quotes a study from 2019 from the Pew Research Center, which found 56% of public law enforcement agencies use facial recognition technology responsibly, and praised the technology’s ability to identify suspects in cases like the Boston Marathon bombing and the riots in the U.S. Capitol earlier this year.
However, it is emphasized that facial recognition is still struggling to accurately identify people who are not white men or East Asian men.
“For any other group, including Native American, black and female groups, (they) still lacks the accuracy required to support the city’s efforts,” it says.
The authors point to major cities like Seattle, Portland, and Oakland that have either banned facial recognition technology or have strict regulations restricting its use.
If Long Beach does not adequately protect against the potential prejudice and risk, facial recognition could “potentially erode public trust in the police,” the report said.
The full one Technology and Innovation Commission is scheduled to discuss the report on Wednesday at 3:30 p.m.