America's Surveillance Scarecrows: AI Camera Towers Are Spreading Fast

What They Are

Across the country, a new fixture is appearing in mall parking lots, residential streets, and public spaces: a compact tow-trailer mounted with a solar panel, battery system, and a telescoping CCTV mast that can extend high above street level. In law enforcement circles, they go by "COWs" — cameras on wheels. On the street, they've picked up a different name: scarecrows.

The devices are sold by companies such as Flock Safety and rented out by large security contractors like Allied Universal. They are designed to fill gaps in static camera networks, deployable anywhere a cellular or WiFi signal exists. Once positioned, they can be integrated directly into local police department feeds.

According to Nile Coates, vice president of US sales at surveillance firm ECAM, the systems are marketed primarily as deterrents. "Our first line of defense is deterrence. We stop crimes before they start," Coates told KTLA5. "This presence alone reduces risk, and when activity escalates, our team can dispatch directly to local guard partners as well as law enforcement," he said.

ECAM alone operates a network of more than 150,000 cameras, Coates told the television station. Logan Harris, CEO of military surveillance company Spotter Global, acknowledged the systems' intimidating appearance. "Sometimes they're referred to as scarecrows because they have bright flashing lights on them, cameras, and they look kind of scary," Harris told KTLA. He added: "The market has spoken. It's been quite amazing how fast this whole market segment has grown. Having that type of video evidence or other sensor data is really helpful."

The Size of the Market

The deployment of COWs is occurring within a much larger expansion of policing technology. According to market research firm Reanin, the police and law enforcement equipment market was valued at $11,723.51 million (approximately $11.7 billion) in 2025, and is projected to grow to $20,288.19 million by 2032, at a compound annual growth rate of 8.2 percent. Reanin attributes the growth in part to increased adoption of "smart wearables, drones, and advanced surveillance systems" across approximately 40 percent of law enforcement bodies.

The mobile surveillance trailer is one of the fastest-growing segments within that market. The combination of declining hardware costs, cellular connectivity, and AI image processing has made it economical for mid-size and small police departments — not just major urban forces — to deploy persistent visual coverage.

AI Facial Recognition on Wheels

The capabilities of modern COW platforms extend well beyond recording video. According to reporting from Futurism, these systems can plug into AI-powered facial recognition pipelines via cellular networks, enabling real-time identity checks against law enforcement databases. The integration turns any street corner into a potential facial recognition checkpoint — without any fixed infrastructure investment by the deploying agency.

Research conducted by YouTuber and security researcher Benn Jordan, in partnership with the online outlet 404 Media, found significant security vulnerabilities in Flock Safety's camera deployments. Jordan found that at least 60 of Flock's AI-enabled "Condor" camera systems had been left openly accessible on the internet, with no username or password protection. The exposed systems allowed anyone to view live video feeds, download up to a month of recordings, and access system settings. Jordan described the high-definition footage — capable of capturing cracks in sidewalks, dents in vehicles, and details of pedestrian clothing — as "Netflix for stalkers," according to WBUR's On Point program.

Flock Safety did not immediately respond to a request for comment from 404 Media at the time of the investigation.

The Federal Layer: Mobile Fortify

The proliferation of private and local COW networks is running in parallel with a federal expansion of mobile facial recognition. The Department of Homeland Security has deployed an app called Mobile Fortify, which allows agents to photograph a person's face in the field and immediately query DHS databases — including passport records, visa files, and border entry photographs — for matches. The system was developed using technology from NEC, a Japanese multinational corporation, according to DHS's 2025 AI Use Case Inventory.

DHS did not publicly acknowledge Mobile Fortify's existence until January 2026, when the agency listed it in its AI Use Case Inventory. By that point, according to court filings cited by Mother Jones, the app had already been used more than 100,000 times since its launch in May 2025. Court filings in a lawsuit filed by the state of Illinois and city of Chicago against DHS revealed that the agency retains all biometric data collected through Mobile Fortify — including data from US citizens — for 15 years.

In a written statement, DHS said: "Mobile Fortify is a lawful law-enforcement tool developed under the Trump Administration to support accurate identity and immigration-status verification during enforcement operations."

The app gained public attention following an incident in Minneapolis in January 2026, when a Somali American US citizen named Abdikafi Abdurahman Abdullahi — known as Kafi — was approached by roughly a dozen ICE agents in an airport rideshare lot and subjected to a facial scan. Abdullahi, who holds a mechanical engineering degree from Washington State University and became a US citizen in 2016, captured the confrontation on video. According to Abdullahi's account in Mother Jones, Mobile Fortify misidentified him, returning a profile for someone named "Ali." DHS declined to comment on Abdullahi's claim that he was misidentified or racially profiled.

Facial recognition technology has a documented history of higher error rates when identifying people with darker skin tones. DHS has not publicly disclosed Mobile Fortify's error rate.

Legal Challenges and Civil Liberties Concerns

The ACLU has filed a class-action lawsuit, Hussen v. Noem, targeting what it describes as a pattern of suspicionless stops, warrantless arrests, and face scans based on perceived race and ethnicity. Nathan Freed Wessler, an attorney with the American Civil Liberties Union, told Mother Jones: "What we've never seen before this year is a law enforcement agency putting face recognition technology on law enforcement agents' phones out in the community and giving them unchecked power to stop people, pull them off the street, and start scanning their faces. It is deeply dangerous. It's irresponsible, it's unprecedented, and it's illegal."

Wessler also noted that advocates have observed surveillance tools being deployed not only to verify immigration status but also to identify and track people who document ICE activity or criticize the agency. "Documenting ICE activity and protesting against it is a right protected by the First Amendment," Wessler said. "Retaliation for doing so goes against the Constitution."

The legal framework governing fixed and mobile surveillance cameras varies dramatically by jurisdiction. Most states have not enacted statutes specifically regulating COW deployments on public property. Federal oversight of AI-enabled facial recognition by state and local law enforcement remains limited.

The Trajectory

In the United Kingdom, the government has pursued a more structured approach. The UK Home Office proposed a national facial matching service for police and a dedicated biometrics regulator in January 2026, with testing slated to begin that year. According to the UK government's own factsheet, live facial recognition deployments by London police between January 2024 and September 2025 led to more than 1,300 arrests for crimes including rape, domestic abuse, and robbery.

In the United States, by contrast, the expansion has largely been market-driven, with police departments independently contracting with private vendors. The combination of mobile COW platforms, AI image processing, and federal tools like Mobile Fortify suggests that the infrastructure for near-continuous public surveillance is already largely in place — and growing — without a corresponding national legal framework to govern it.

Reanin projects the US law enforcement equipment market will approach $20.3 billion in value by 2032. Much of that growth is expected to come from surveillance and connected technologies, not traditional gear.