What is AI surveillance in commercial real estate? AI surveillance in commercial real estate is the use of cameras, automated license plate readers, and access control platforms that apply computer vision to identify vehicles, people, and events across a property, then route alerts to staff or law enforcement. On September 21, 2026, Bisnow reported this market is expanding quickly among private landlords even as cities cancel contracts in large numbers. That divergence is the story: the liability municipalities are shedding is landing on commercial property. For the wider picture, see our guide to AI property management.
Key Takeaways
- More than 200 US cities and counties have ended Flock Safety relationships since 2021, roughly 90 of them in August 2026 alone, even as private landlords expand deployment.
- Flock reached roughly $500 million in annual recurring revenue by March 2026, and research firm Sacra estimates about half of new revenue now comes from business customers.
- Simon Property Group, the largest US mall owner, faces a February 2026 class action over license plate readers at 23 California properties.
- California Civil Code Section 1798.90.51 binds private ALPR operators, not just police, requiring a conspicuously posted privacy policy covering seven items.
- Section 1798.90.52 requires a query-level access log naming who searched, what plate, and why. Its absence is a central allegation against Simon.
- At roughly $2,400 per camera per year, a 20 camera center carries about $48,000 of annual expense, or roughly $738,000 of value at a 6.5% cap rate if nothing offsets it.
AI Surveillance in Commercial Real Estate Is Migrating From Public to Private Buyers
The short answer: municipal demand is collapsing and enterprise demand is replacing it. Published tallies of council votes show more than 200 US cities and counties have ended Flock Safety relationships since 2021, roughly 90 in August 2026 alone, with a fresh wave across Indiana counties in September. Florida's Department of Transportation revoked highway ALPR permits that same month. The trigger has been data sharing: in August 2026, Flock acknowledged to Denver's 9News a pilot with US Customs and Border Protection after CEO Garrett Langley had denied federal contracts.
None of that has slowed the private side. Bisnow's September 21 reporting describes a full vendor ecosystem selling AI video to landlords and occupiers. Brivo reports tens of thousands of customers across more than 80 countries, with Greystar, Whole Foods, and Prada among named clients, and Verkada had roughly 30,000 customers at the end of 2025. Industry projections cited in that coverage put the AI video surveillance market near $6.3 billion in 2025, growing toward $27 billion by 2034. Marcelo Ballve of the research firm Sacra estimates about half of Flock's new revenue now comes from businesses. Brivo CEO Steve Van Till framed the appeal plainly: most customers are private entities, and they keep the data for themselves. That is exactly the feature that makes private deployment attractive and the exposure that makes it risky.
Why the Private Lane Has Less Oversight, Not Less Risk
Lighter oversight is not lighter liability. California public agencies must hold a public meeting before switching on an ALPR system and generally cannot share data outside government. Private operators face no such requirement, which is why the vendor pipeline rotated toward them. What they do face is a private right of action. The operational case is still real: our breakdown of AI security and access control for commercial properties covers where these systems cut guard hours, and JLL's Smart Building Essentials guide helps decide what to scale. What changed in 2026 is that data governance became a litigation exposure.
The Simon Property Group Class Action Is the Warning Shot
The most instructive event for CRE owners is a lawsuit, not a product launch. On February 17, 2026, plaintiffs' firm Bursor & Fisher filed Leonard v. Simon Media Properties, LLC in San Diego County Superior Court, targeting license plate reader deployment across Simon Property Group's 23 California properties. Reporting indicates it was one of four ALPR class actions brought within about six weeks of a February 2026 California appellate ruling on license plate data.
The allegations read as a checklist of what not to do. The complaint contends Simon added ALPR language to its privacy policy in February 2026 but buried it behind multiple hyperlinks, requiring three clicks to reach what the complaint called the opposite of conspicuous. Plaintiffs also target open law enforcement access without audit records, sharing with out-of-state and federal agencies, undisclosed participation in Flock's Business Network, and missing accuracy safeguards. These are allegations, not findings, and the case remains pending.
The pattern generalizes. Any owner running plate readers in a parking field, and that includes most large retail centers and a growing share of multifamily communities, operates the same technology under the same statute.
What California Law Actually Requires of Private Landlords
California Civil Code Section 1798.90.51 applies to any ALPR operator, defined by the technology in use rather than by whether the operator is a government body. If you run the cameras, you are the operator. The statute requires reasonable security safeguards plus a usage and privacy policy posted conspicuously on your website, addressing seven items at minimum:
- Authorized purposes for operating the system and collecting data.
- Authorized users by job title, including contractors, with required training.
- Monitoring for security and privacy-law compliance.
- Sharing rules governing any sale, sharing, or transfer to other persons.
- Named custodian, by title, responsible for implementation.
- Accuracy measures to keep data correct and fix errors.
- Retention and destruction periods and the process for deciding when to delete.
Section 1798.90.52 adds the requirement that trips up most private deployments. If you access the data or provide access to anyone else, you must record each access: date and time, the plate queried, the username and affiliation of the person querying, and the purpose. A standing login handed to a police department with no query log does not satisfy that. Section 1798.90.53 imposes parallel duties on end users and adds periodic audits. The statutory text sits at the California Legislative Information portal. Owners elsewhere should not relax; this is the template other states draw from, and the same themes run through our coverage of AI data governance rulings.
How to Underwrite the Cost and the Exposure
Start with the line item. Flock's published model runs about $2,400 per camera per year plus a roughly $350 installation fee, so a 20 camera retail center carries about $48,000 of annual operating expense. Because NOI is gross revenue less operating expenses, that reduces NOI dollar for dollar unless something offsets it, and at a 6.5% cap rate it maps to roughly $738,000 of value, since value equals NOI divided by the cap rate. Weigh that against the loss prevention and insurance benefits, a tradeoff The AI Consulting Network helps owners model before a contract gets signed. Then price the tail: statutory privacy claims attract plaintiffs' firms because damages can be assessed per person across a class, and a grocery-anchored center processes an enormous number of vehicles. Before signing or renewing, run five checks:
- Read your own privacy policy and count the clicks from the homepage. If the ALPR disclosure is not conspicuous, fix it first.
- Audit third party access. Identify every agency or business with a login and confirm a query-level log exists.
- Decide on network sharing deliberately. Opting into a vendor's cross-customer network is a sharing decision your policy must disclose.
- Set a retention window and enforce it. Indefinite retention is the hardest position to defend.
- Push indemnity to the vendor and confirm your liability and cyber policies respond to statutory privacy claims.
The same governance discipline applies to every AI system touching sensitive data, which we cover in our guide to AI model security and data privacy for CRE investors. Investors who want hands-on help building an AI vendor diligence process can reach out to Avi Hacker, J.D. at The AI Consulting Network.
Frequently Asked Questions
Q: Does California's ALPR law really apply to private landlords?
A: Yes. Section 1798.90.51 defines an ALPR operator by the technology being run, not by whether the operator is a public agency. A shopping center, industrial park, or apartment community running plate readers carries the same policy, security, and logging duties, and the statute provides a private right of action.
Q: Why are cities canceling Flock contracts if landlords are expanding?
A: Cities answer to voters at public meetings, and federal data sharing became untenable after Flock acknowledged a Customs and Border Protection pilot in August 2026. Private landlords face no equivalent vote, so vendors rotated toward enterprise buyers.
Q: What is the single most common compliance gap?
A: Access logging. Many owners grant a police department a standing login as a goodwill gesture and keep no record of individual queries. Section 1798.90.52 requires a per-query record of who searched, what plate, and why. Its absence is a core allegation in the Simon litigation.
Q: How does this affect property valuation?
A: Through NOI and through risk. Camera subscriptions reduce NOI directly, and at a 6.5% cap rate every $48,000 of unoffset annual expense corresponds to roughly $738,000 of value. Litigation exposure is harder to price but increasingly surfaces in buyer diligence. For personalized guidance on evaluating these systems, connect with The AI Consulting Network.