AI Weekly: New Poll Shows Public Opinion on Facial Recognition, DOJ Doesn’t Track Predictive Police Spending


Missed a session at the Data Summit? View on demand here.

This week in AI, a new Pew Center survey shed light on Americans’ view of AI, including police’s use of facial recognition. In other news, the US Department of Justice revealed it has not kept any “specific data”[s]about his purchases of predictive policing technologies, a category of technologies that investigations have found to be biased against minority groups.

Lured by the promise of crime reduction and time to solve cases, law enforcement agencies have increasingly explored AI-powered tools, such as facial recognition, drones and predictive police software, that attempt to predict where crime will occur using historical data. According to Markets and Markets, law enforcement agencies are expected to spend as much as $18.1 billion on software tools, including AI-powered systems, up from $11.6 billion in 2019.

But the effectiveness of these systems has been repeatedly questioned. For example, an investigation by the Associated Press found that ShotSpotter, a “gunfire location service” that uses AI to triangulate the source of firearms discharges, can miss live gunshots directly under its microphones or misclassify the sounds of fireworks or backfires. Extensive reporting by Gizmodo and The Markeup, meanwhile, has revealed that Geolitica (formerly called PredPol), a police software that attempts to anticipate property crimes, disproportionately predicts that crime will be committed in neighborhoods inhabited by working-class people, people of color, and black people in particular.

Facial recognition has also been shown to be biased against ‘suspects’ of certain skin colors and ethnicities. At least three people in the US – all black men – have been wrongfully arrested on the basis of poor facial recognition matches. And studies, including the groundbreaking Gender Shades project, have shown that facial recognition technology once sold to law enforcement, including Amazon’s Rekognition, is significantly more likely to misclassify the faces of dark-skinned people.

But dichotomously, public support for police use of facial recognition is relatively high, with many respondents to a recent Pew report saying they agree with its deployment. The reason could be the relentless PR campaigns of vendors like Amazon, who have claimed that facial recognition can be a valuable tool in helping find missing persons, for example. Or maybe it’s ignorance of the technology’s shortcomings. According to Pew, respondents who heard a lot about the use of facial recognition by the police were more likely to say it’s a bad idea for society than those who hadn’t heard about it.

Racial divisions surfaced in the results of the Pew survey, with black and Hispanic adults more likely than white adults to say that police would definitely or probably use facial recognition to check black and Hispanic neighborhoods more often than other neighborhoods. Given that black and Hispanic individuals are more likely to be arrested and incarcerated for petty crimes and are consequently overrepresented in mugshot data — the data historically used to develop facial recognition algorithms — it isn’t. surprising.

“Remarkable areas of people’s lives are now being tracked and monitored by police, government agencies, businesses and advertisers… Facial recognition technology adds an extra dimension to this problem as all kinds of surveillance cameras can be used to pick up details about what people are doing in public places and sometimes in stores,” the Pew study co-authors write.

Predictive policing of the Ministry of Justice

The Department of Justice (DOJ) is a growing investor in AI and has awarded Veritone a contract to provide transcription services for its attorneys. The department is also a customer of Clearview, a controversial facial recognition provider, where employees of the FBI, the Drug Enforcement Administration and other DOJ agencies have used it to conduct thousands of searches for suspects.

But according to Gizmodo, the DOJ keeps poor records of its spending, especially when it comes to predictive police tools. Speaking to the publication, a senior official said the Department of Justice does not actively monitor whether funds from the Edward Byrne Memorial Justice Assistance Grant Program (JAG), a leading source of criminal justice funding, are being used to purchase predictive police services.

That’s alarming, say Democratic senators, including Ron Wyden (D-OR), who sent a letter to US Attorney General Merrick Garland in April 2020 requesting basic information about the DOJ’s funding of AI-powered software. Wyden and his colleagues expressed concern that this software lacked meaningful oversight, may have heightened racial bias in the police force, and may even violate citizens’ right to a fair trial.

The fears are not unfounded. Gizmodo notes that audits of predictive tools “have found no evidence that they are effective in preventing crime” and are often used “without transparency or… opportunities for public input.”

In 2019, the Los Angeles Police Department, which had trialled a range of AI policing tools, acknowledged in an internal review that the tools “often deviated from their stated goals.” In a study that same year, researchers at New York University found that nine police departments had entered software data generated “during periods when the department had engaged in various forms of unlawful and biased police practices.

“It’s unfortunate that the Department of Justice has chosen not to answer the majority of my questions about federal funding for predictive police programs,” Wyden said, suggesting to Gizmodo that it may be time for Congress to ban the police. weighing technology. A number of cities, including Santa Cruz, California and New Orleans, Louisiana, have already banned the use of predictive police programs. But partisan stalemate and special interests have hampered efforts at the federal level so far.

For AI reporting, send news tips to Kyle Wiggers – and subscribe to the AI ​​Weekly newsletter and bookmark our AI channel, The Machine.

Thank you for reading,

Kyle Wiggers

Senior AI Staff Writer

VentureBeat’s mission is to be a digital city square for tech decision makers to learn about transformative business technology and transactions. Learn more

This post AI Weekly: New Poll Shows Public Opinion on Facial Recognition, DOJ Doesn’t Track Predictive Police Spending

was original published at “”