Cops pause use of flawed AI cameras secretly monitoring streets for suspects
New Orleans police have reportedly spent years scanning live feeds of city streets and secretly using facial recognition to identify suspects in real time—in seeming defiance of a city ordinance designed to prevent false arrests and protect citizens’ civil rights.
A Washington Post investigation uncovered the dodgy practice, which relied on a private network of more than 200 cameras to automatically ping cops’ phones when a possible match for a suspect was detected. Court records and public data suggest that these cameras “played a role in dozens of arrests,” the Post found, but most uses were never disclosed in police reports.
That seems like a problem, the Post reported, since a 2022 city council ordinance required much more oversight for the tech. Rather than instantly detaining supposed suspects the second they pop up on live feeds, cops were only supposed to use the tech to find “specific suspects in their investigations of violent crimes,” the Post reported. And in those limited cases, the cops were supposed to send images to a “fusion center,” where at least two examiners “trained in identifying faces” using AI software had to agree on alleged matches before cops approached suspects.
Related Knowledge Center: Artificial intelligence
On top of a failing business model, newsrooms must navigate the powerful — and not fully reliable — new capabilities of artificial intelligence and the deployment of large language models.
State of the business model | Copyright and intellectual property
Related Knowledge Center: State surveillance
While government monitoring of personal digital communications has traditionally been associated with authoritarian regimes such as China, the Trump regime has moved to use citizens’ personal information in unprecedented ways.
Digital privacy and data security | FISA — Section 702