Look up between the Union and the Cathedral.

Signs at Pitt’s cameras say the area is “subject to 24 hour monitoring.” The same policy says the cameras aren’t watched under normal conditions, only for listed reasons such as alarms, restricted areas and investigations.¹⁰ Pitt Police and ISD both monitor the system.¹²

Footage can be released only with approval from the Senior Vice Chancellor for Business and Operations, the Associate Vice Chancellor for Public Safety and Emergency Management, the Chief of Police or University Counsel.¹⁰ In 2022 Pitt’s vice chancellor for public safety said someone who wants footage of a crash on Forbes for a lawsuit needs a subpoena. He also gave an example of what the footage can do: after a hit-and-run on Fifth Avenue, Pitt Police got the driver’s plate and tracked the driver down.²

ISD buys all surveillance equipment for the University, and the Department of Public Safety says it will follow new developments in the law and in the security industry.¹⁰

Face recognition has contributed to more than a dozen known wrongful arrests, and nearly everyone misidentified has been Black; the technology is measurably less accurate for people with darker skin.^13,14 Documented cases include a man handcuffed in front of his children, a woman questioned while eight months pregnant, and a man jailed over a holiday weekend for a crime in a state he had never visited.¹³

A recurring step compounds the error: police build a photo lineup around the software’s pick, so a witness can end up confirming the same face the software chose. Departments say a match is only a lead, and some, like Detroit, have tightened their rules. In the documented cases, the match was treated as more than a lead.¹³

Oakland, California publishes its numbers. In 2025 its 293 readers logged 638,747,333 plate scans and 1,099,837 “hotlist” alerts, about one alert for every 580 scans, and only a fraction of the alerts led anywhere.¹⁹ More than 99.8 percent of the scans matched nothing on a hotlist.

Records the Electronic Frontier Foundation obtained show more than 12 million searches of Flock’s network by over 3,900 agencies in under a year, including searches tied to political protests, searches targeting Romani people, and searches related to reproductive healthcare.¹⁶ Reporters documented Texas officers searching plate data for a woman who had traveled for an abortion, and federal immigration authorities reaching one state’s data in apparent violation of its privacy law, which led to a state audit.^16,20 Officers have also used these systems to track specific people, including a case on our advocacy page.

Across late 2024 and 2025, independent researchers catalogued security flaws in Flock’s devices. One white paper documented 51 findings, 22 of them assigned formal vulnerability IDs.^21,22 Researchers showed that brief physical access to a roadside camera was enough to take it over, and reported older units running outdated software with unencrypted storage.²² Dozens of cameras were found online with no password, showing live feeds of parking lots, arguments and children in parks, many of them cameras that track people rather than plates.²³

Flock says it registered the vulnerabilities with the national CVE database and that none of them undermine its customers’ public-safety work.²¹ It first argued the attacks needed physical access and inside knowledge; researchers replied that mounted, cloud-connected cameras showed the same weaknesses.²² Lawmakers have asked the Federal Trade Commission to investigate, citing stolen customer credentials and no required multi-factor authentication.¹⁶

Flock’s own analysis estimates its network helps solve about 700,000 crimes a year, roughly a tenth of reported crime nationwide.²⁴ Police point to cases like a suspect identified in the killings of three women, and an Oakland NAACP chapter argues an automated read is less biased than an officer’s choice of whom to stop.²⁵

Independent evidence is thinner. The Electronic Frontier Foundation, which opposes these networks, says there is little empirical evidence that they reduce crime.²⁶ A study of a large expansion in Atlantic City found no clear drop in violent crime overall, with some links to fewer shootings and thefts that the authors said other factors could explain.²⁰ Oakland’s police call the cameras “one element,” not the cause of any decline.¹⁹ “This camera helped solve a case” and “recording everyone is justified” are different claims, and the evidence above supports the first far better than the second.

Flock also sells acoustic sensors first marketed for detecting gunshots. In October 2025 it said it would expand them to listen for “human distress,” including screaming, and camera drones are increasingly sent out as first responders.¹⁶

The Federal Trade Commission charged that Ring let employees and contractors browse customers’ private videos, and that its security was weak enough for hackers to take over cameras. By the FTC’s account, one employee spent months viewing thousands of recordings from women’s bedrooms and bathrooms before a colleague noticed. Ring settled for $5.8 million, was ordered to delete algorithms built from improperly viewed footage, and must require multi-factor authentication.³³

The Electronic Frontier Foundation’s concern with Ring is police getting doorbell footage without a warrant.³⁴

The vacuum attack could also delete the spoken “camera on” warning. The researchers said Ecovacs took about ten months to respond; the company later blamed real-world incidents on recycled passwords and promised a fix.³⁵

Vizio was ordered to get opt-in consent going forward. It noted the viewing data wasn’t tied to a name; it was tied to the household’s IP address.³⁷

The Copyright Office grants narrow repair exemptions every few years, but the default runs the other way, which is why the right-to-repair fight and the surveillance fight keep turning out to be the same fight.³⁸ Cars follow the same pattern; see what your car logs.