Former Pentagon AI chief warns Flock cameras could create 'China-like Ai surveillance state' in the US
- Gammatek ISPL
- 16 hours ago
- 5 min read
By Gammatek ISPL Last updated: September 2026 | 13 min read
Why This Matters to You
If you've driven past a black camera pole with a solar panel on top anywhere in the US in the last few years, there's a good chance it logged your license plate, the time, and your direction of travel — and stored that information in a searchable database accessible to police departments, and in some documented cases, federal agencies including ICE, the Secret Service, and even the Navy's criminal investigation division. This isn't a hypothetical future risk. It's already operating in more than 6,000 communities, run primarily by a single company, Flock Safety, whose network now scans an estimated 20 billion vehicle sightings a month. Whether you think that's a reasonable trade-off for public safety or a dangerous expansion of unaccountable tracking, the infrastructure is already built and expanding — which is exactly why it matters to understand what it actually does before deciding how you feel about it. https://link.amazon/B0aHnrrgF
What Flock Safety's System Actually Does
Founded in 2017, Flock Safety builds automated license plate reader (ALPR) cameras that photograph passing vehicles and use AI to log the make, model, color, and license plate into a central database. The company has grown into a roughly $7.5 billion business, selling primarily to police departments, homeowners associations, and private businesses.
Flock has publicly stated its cameras capture vehicle information only, not biometric data on individuals. But independent reporting has found the system's capabilities extend further: it can identify and track a vehicle's "associates" based on movement patterns, and when combined with existing law enforcement databases, can be used to infer a person's home address, routines, and social connections — even without directly capturing a face or fingerprint. https://link.amazon/B05COJgC6
The Bipartisan Alarm
What makes this story notable isn't just the technology — it's who's raising concerns about it, from very different political directions.
Senator Bernie Sanders (I-VT) has been among the most vocal critics, pledging to introduce federal legislation to "stop Flock and AI mass surveillance." He's described the company as backed by prominent tech investors and warned that its 120,000-plus cameras risk pushing the country toward what he called a surveillance-driven erosion of constitutional privacy rights.
Senator Josh Hawley (R-MO), chair of the Senate Judiciary Committee's crime and counterterrorism subcommittee, launched a formal investigation into Flock's practices — a notable move from a lawmaker not typically aligned with Sanders on much else. Hawley's letter to Flock's CEO argued that the majority of people captured in the company's database "had done nothing wrong."
Senator Ron Wyden (D-OR) has separately raised alarms, warning that certain federal data practices involving AI-driven vehicle tracking represent what he called a "chilling expansion of mass surveillance."
Representative Thomas Massie (R-KY) has reportedly moved to introduce a funding-cutoff bill targeting the technology's federal use.
This kind of left-right alignment on a single tech policy issue is rare, and it's worth paying attention to precisely because it doesn't fit the usual partisan pattern — when lawmakers who agree on almost nothing else converge on the same concern, it's a signal the underlying issue isn't just political theater.
Real Documented Problems, Not Just Hypotheticals
Critics of the "surveillance state" framing sometimes argue it's alarmist — that these are just cameras helping solve real crimes. There's truth to that: law enforcement agencies have credited the system with recovering stolen vehicles and advancing investigations, including reportedly helping locate a suspect in a high-profile shooting case.
But documented misuse cases complicate the purely positive framing:
Reports have surfaced of the technology being used to falsely accuse people of crimes based on flawed matches.
Investigations found that law enforcement officers in some jurisdictions accessed the system for personal, non-official reasons.
A "National Lookup" tool reportedly enabled out-of-state searches that bypassed local legal restrictions on immigration enforcement data-sharing, prompting several cities to cancel contracts or restrict federal access entirely.
Independent security researchers have separately raised concerns about weak security controls in the underlying camera hardware and software, given how widely deployed the system now is.
Flock has responded to some of this criticism by restricting federal agency access to certain lookup tools and stating it does not directly partner with immigration enforcement — though it acknowledges individual local police departments control who they choose to share data with.
Company's stated position | Independent findings | |
Data captured | Vehicle information only | Can infer associates, addresses, routines via cross-referencing |
Federal agency access | Barred from national/state lookup tools (as of last year) | Multiple federal agencies previously had access, per Senate letter |
Accuracy | System aids investigations | Documented cases of false accusations from flawed matches |
Misuse | Not the company's responsibility once sold to local police | Documented cases of personal/unauthorized use by officers |
The "China-Like" Comparison — Fair or Overstated?
Comparisons to China's state surveillance apparatus are common in this debate, and they deserve scrutiny rather than uncritical repetition in either direction.
The case for the comparison: China's surveillance state relies on the same basic architecture — dense camera networks, AI-driven identification, centralized data access for state authorities — deployed at a scale that normalizes constant tracking as background infrastructure. Critics argue that the technical capability gap between the two systems is narrower than most Americans assume, even if the current legal and political context differs.
The case against treating it as equivalent: China's system operates within a one-party state with no independent judiciary, no free press to investigate abuses, and explicit government mandates for population control and dissent suppression — including documented use against ethnic and religious minorities. The US system, however flawed, currently operates within a framework of Congressional oversight (as this very controversy demonstrates), a free press actively reporting on abuses, and local governments that have in some cases successfully canceled contracts or imposed restrictions.
The honest position is probably: the technology itself doesn't inherently distinguish between a democracy and an authoritarian state — the safeguards, oversight, and accountability structures around it do. That's precisely why the current Congressional scrutiny matters: it's one of the mechanisms that differentiates "a country with surveillance technology" from "a surveillance state," and its outcome will say a lot about which direction this settles into.
An Implementation Consideration for Organizations Evaluating Similar Technology
For any organization — a business, a school district, a homeowners association — considering deploying AI-powered camera or monitoring systems, this controversy offers a concrete lesson: the technical capability of a system tends to expand faster than the policy governing its use. Organizations adopting this kind of technology should establish, in writing, before deployment:
Exactly who can access captured data, and under what circumstances
Data retention limits and deletion schedules
Whether and how the data can be shared with third parties or other agencies
An audit trail for every access request, reviewable by an independent party
Systems deployed without these guardrails in place from day one tend to have them added reactively, after a controversy — which is a much harder position to recover from, reputationally, than building them in from the start.
Where This Goes Next
Congressional investigations, proposed legislation, and local contract cancellations are all still in progress as of this writing — meaning the regulatory outcome here is genuinely unresolved. Whether this results in federal restrictions, industry self-regulation, or continued expansion largely unchanged will likely become clearer over the next several months as Hawley's investigation and Sanders' proposed legislation move forward (or don't).
What's already clear is that the technical capability for widescale, AI-driven movement tracking exists today, is commercially available, and is already deployed at a meaningful scale — the debate now is entirely about what oversight, if any, gets built around it. https://www.gammateksolutions.com/post/america-must-learn-ai-lessons-from-astro-boy
Comments