The Watchful Eye's Blind Spot: Flock Safety Cameras Under Fire for Rampant Errors

Imagine a network of watchful eyes across your city, tirelessly scanning license plates, promising to catch criminals and keep communities safe. That's the vision Flock Safety, a prominent Atlanta-based company, sells with its automatic license plate recognition (ALPR) cameras. With over 120,000 cameras deployed in nearly every state, serving thousands of law enforcement agencies and communities, their presence is undeniable. The company proudly touts impressive accuracy rates, claiming over 99% capture rates and 93% accuracy for plate readings, with billions of scans happening monthly. Sounds reassuring, right?

But here's where the vision gets blurry, or rather, alarmingly distorted. A deep dive by Business Insider into data from the Roseville Police Department in California uncovered a troubling reality. Over a two-year period, from 2023 to 2024, Roseville PD analyzed 1,427 alerts generated by Flock cameras, specifically those flagged for stolen vehicles or felony crimes. The findings? A staggering 71% of those alerts were false positives. Yes, you read that correctly – nearly three-quarters of the time, the cameras got it wrong, flagging innocent plates as suspicious. Flock, for its part, tried to brush this off, citing 'unique deployment' issues in Roseville, like older cameras and unusual installation heights. They even suggested their performance has 'significantly improved' since then. Yet, Roseville PD, having already sunk around $450,000 into the system and communicated these issues to Flock for four years, strongly disputes any claims of improvement. They haven't made a single arrest based solely on a faulty Flock scan. That's a pretty strong indicator, isn't it?

The Roseville case isn't an isolated hiccup; it's a symptom of a much larger problem, one that has devastating real-world consequences for innocent people. The Institute for Justice, a non-profit advocating for individual liberties, has meticulously documented at least 27 cases nationwide where drivers were mistakenly pulled over, detained, or even arrested because a Flock camera simply misread a license plate. And let's be clear, 27 is likely just the tip of the iceberg. Imagine being an innocent couple in Sherwood, Arkansas, pulled over at gunpoint by officers in February because a camera misidentified your SUV. Or consider the harrowing experience of Mr. Parra in San Diego, wrongly jailed for almost a month in November 2025. His car was falsely linked to an attempted carjacking by a Flock camera, even though he was five miles away from the crime scene. Another driver in Cherry Hills, Colorado, found himself repeatedly stopped, his plate mistakenly stuck on a police hotlist. These aren't just statistics; they're stories of lives upended by technology that's supposed to make us safer.

It truly begs the question: if these systems are so error-prone, are they actually helping? In Oak Park, Illinois, perhaps the only jurisdiction brave enough to regularly publish ALPR error reports, the picture isn't much better. Their data consistently shows that a third or more of all traffic stops prompted by a Flock alert end with the driver being released because of data inaccuracies. An oversight board in Oak Park even concluded in 2025 that Flock's ALPRs haven't played a 'meaningful role' in any crime investigation since their installation back in 2022. That's a serious indictment of a system marketed as a crime-fighting powerhouse.

Beyond just misidentification, there's an even darker side: the potential for misuse and the chilling erosion of privacy. A CNN exposé, highlighted by the Institute for Justice, revealed a disturbing pattern: dozens of law enforcement officers have reportedly abused Flock data, using it to stalk romantic interests or former partners. An IJ survey identified 28 such cases, with two recent incidents in Texas in July 2026, including an officer stalking a female colleague. It's truly shocking. Flock Safety, to their credit, has introduced a new AI auditing tool to try and catch these abuses, and it seems to be having some effect – at least 20 officers in Georgia have faced discipline or charges recently. A Flock spokesperson, while calling misuse 'very rare,' admitted it's 'completely unacceptable.' And it's not just personal stalking; federal immigration agents are also reportedly leveraging Flock data to track and detain undocumented immigrants, adding another layer of controversy. This is why many cities, like Austin and Bandera in Texas, have chosen to cut ties with Flock, terminating contracts over profound privacy concerns. We're talking about 47 cities nationwide making this tough decision. It really makes you wonder about the trade-offs we're making for perceived security.

So, we're left with a significant disconnect. On one hand, Flock Safety paints a picture of nearly infallible technology, a digital guardian against crime. On the other, we see mounting evidence from real police departments and victim accounts suggesting a system riddled with errors, prone to misuse, and often failing to deliver on its grand promises. The technology itself isn't inherently bad, perhaps, but its current implementation, accuracy, and oversight clearly leave much to be desired. As these pervasive networks of cameras continue to expand, it becomes increasingly crucial to demand greater transparency, stricter accountability, and a serious reevaluation of whether the trade-off between surveillance and genuine public safety is truly worth it – especially when innocent lives are repeatedly caught in the crosshairs of mistaken identity.