The 120,000-Camera Network That Outruns the Fourth Amendment: Sanders vs. Flock's Surveillance-as-a-Service

Bernie Sanders just declared war on a company most Americans have never heard of. The target: Flock Safety, which has deployed over 120,000 AI-powered cameras across the United States. But here's the data point that should stop you cold: this isn't a technology story. It's a network effects story. And the network is already larger than anything the federal government has built for itself. Sanders' promise to introduce legislation blocking Flock's mass surveillance marks a significant escalation. The Vermont senator framed the issue in stark terms: Flock's network is a path to a "surveillance state." The framing is politically charged, but the underlying mechanics deserve forensic attention. Because when you strip away the rhetoric, what Flock has built is a privatized, subscription-based national vehicle trajectory database - and the Fourth Amendment has not yet caught up with it. Flock Safety, founded in 2017, has raised over $380 million in venture funding. Its business model is deceptively simple: deploy cameras at low upfront cost - roughly $3,000 per unit - then charge annual subscription fees between $2,000 and $5,000 per camera. The hardware combines automatic license plate recognition (ALPR), vehicle classification via CNN-based image analysis, and acoustic gunshot detection. None of this is frontier AI. The OCR and pattern recognition technology is mature, commercially available, and frankly unremarkable. What is remarkable is the network. Every Flock camera feeds into a shared, searchable database of vehicle trajectories. A camera in suburban Ohio can help identify a vehicle last seen in California. Law enforcement agencies across jurisdictions can query each other's data, creating what is functionally a national surveillance grid without any federal mandate or oversight. Flock's growth trajectory is worth examining. From roughly 1,000 cameras in 2019 to over 120,000 today, the deployment curve resembles the adoption curve of a successful software platform, not a hardware vendor. The company's viral expansion strategy is deliberate: when one community installs Flock cameras, neighboring communities feel pressure to do the same, creating a contagion effect that drives organic growth. This is the same dynamic that made Ring doorbells ubiquitous, but with a critical difference: Flock's cameras are networked and searchable, not isolated consumer devices. The legal landscape is unsettled. The Supreme Court's 2018 Carpenter v. United States decision held that warrantless access to seven or more days of cell phone location data violates the Fourth Amendment. But whether that precedent extends to privately-collected license plate data remains unresolved. Flock's argument - that license plates are public information and photographing them on public roads does not constitute a "search" - has found support in several district courts. Sanders' legislative promise sits within a broader wave of AI surveillance legislation. States including Maine, Utah, and Vermont have passed ALPR restrictions. The Fourth Amendment Is Not For Sale Act has been under Senate consideration since 2023. Bipartisan letters demanding Flock transparency have been sent. But no federal framework exists. The gap between state-level patchwork and federal inaction is precisely where Flock's network has grown. The technical architecture matters more than the political rhetoric. Flock's competitive moat is not algorithmic sophistication - it's deployment density. 120,000 cameras create a data flywheel: more cameras generate more vehicle trajectories, which increases search value, which attracts more customers, which funds more deployments. This is the same network effect dynamic that drives blockchain protocols, and it has the same consequence: the value of the network grows superlinearly with the number of nodes. This is surveillance-as-a-service. The innovation isn't technical; it's financial engineering. By lowering the upfront cost barrier, Flock made it possible for cash-strapped suburban police departments, homeowners associations, and private businesses to build what amounts to a privatized national surveillance infrastructure. The company actively sells to HOAs and private enterprises, meaning the network is not solely under government control. This complicates any regulatory response - you cannot simply restrict government use when private entities control a significant portion of the data collection. The data retention question is where the legal battle will be fought. Default retention is 30 days, but law enforcement can request extensions. Third-party access is granted to non-law enforcement entities. Cross-jurisdictional data sharing is the default, not the exception. And critically, Flock has begun testing facial recognition capabilities in some jurisdictions - a technological escalation that transforms vehicle tracking into person identification. From my perspective as someone who has spent years analyzing on-chain data flows, the parallels are uncomfortable. In blockchain, we celebrate transparency - every transaction is visible, auditable, permanent. Flock's network offers the same permanence, but without the accountability. There is no public ledger of who queries the database, what searches are performed, or how the data is used. The transparency is one-directional: citizens are visible to the system, but the system is opaque to citizens. The network remembers what the law forgets. Surveillance scales geometrically; accountability scales arithmetically. This asymmetry is the core structural problem. Flock's network grows through a viral deployment model - when one community installs cameras, neighboring jurisdictions follow, creating a contagion effect. The company's own marketing emphasizes this: each new camera increases the value of every existing camera. This is precisely the network effect that makes the system so difficult to regulate. You cannot simply remove one node; the value is distributed across the entire graph. The ICE connection is the unspoken elephant. Progressive lawmakers are concerned that local surveillance data could be shared with federal immigration enforcement. This is not hypothetical - the infrastructure for such sharing exists, and the legal barriers are minimal. The chilling effect on immigrant communities is well-documented in academic research on ALPR systems. The concern is not merely theoretical; Flock has already faced scrutiny over its data sharing practices, and the company's opacity regarding its federal partnerships has fueled speculation. The disparate impact is also measurable. Studies have shown that ALPR systems lead to disproportionate stops of minority drivers. The deployment pattern creates a "surveillance divide": affluent communities can afford Flock's protection, while lower-income communities face either over-surveillance or no protection at all. This is not a bug in the business model; it is the business model. The legislative path forward is murky. Sanders could pursue several routes: federal ALPR restrictions, warrant requirements for third-party data access, or extending Electronic Communications Privacy Act protections to ALPR data. Each path faces constitutional and political obstacles. Federalism questions loom - Flock's cameras are primarily deployed by local governments, and federal legislation may face state sovereignty challenges. Law enforcement opposition is guaranteed; Flock's systems have demonstrably helped solve crimes, and police unions will lobby against restrictions. The political timing is also worth noting. Sanders' announcement arrives in an election cycle where AI surveillance is a mobilizing issue for progressive voters and civil liberties advocates. The legislative promise may be as much a political signal as a policy commitment. But even if the bill fails, the framing shift matters: Flock has been redefined from a "local crime-fighting tool" to a "national surveillance threat." That reframing changes the regulatory conversation permanently. Here is the counter-intuitive angle: the AI is not the problem. The cameras are not even the problem. The problem is the business model that externalizes surveillance costs while internalizing data value. Sanders' legislative push, if it focuses on restricting ALPR technology itself, will likely fail. The technology is too embedded, too useful for legitimate law enforcement, and too popular with the suburban communities that purchase it. The more effective regulatory lever is data retention and access control. A law requiring warrants for third-party data access, or capping retention at seven days, would gut the network's value without banning the technology. This is the "correlation is a map, but causation is the terrain" problem: lawmakers are targeting the visible symptom (cameras) rather than the underlying mechanism (the data network). Data retention is policy; deletion is a promise. But here is the uncomfortable truth: even if Sanders succeeds, Flock's competitors - Motorola Solutions, Axon - are waiting to absorb the demand. Restricting one vendor does not dismantle the surveillance economy; it merely reshuffles market share. The regulatory window is narrow, and the industry is already adapting. Flock could preemptively shorten retention periods, restrict facial recognition, and publish transparency reports to defuse legislative momentum. If that happens, the urgency evaporates - and the surveillance infrastructure remains, just with slightly better PR. The signal to watch is not Sanders' rhetoric. It is whether Flock preemptively adjusts its own policies to defuse legislative momentum. If Flock self-regulates, the urgency evaporates. If it does not, the next 12 months will determine whether the Fourth Amendment applies to privately-built surveillance networks. The ledger of American privacy is being written in camera deployments, not court rulings. And unlike a blockchain, that ledger is not publicly auditable. The question is whether we demand transparency before the network becomes irreversible.