What Flock Safety's Marketing Crisis Teaches AI Companies

By July 2026, Flock Safety had the kind of contradiction every growth-stage AI company should study before it becomes their own. The company was still describing a fast-growing business, with Forbes reporting roughly $500 million in annual recurring revenue and an $8.3 billion valuation, while its CEO was publicly apologizing for having called DeFlock, an activist mapping project, a “terroristic organization” months earlier.[1] At the same time, critics were counting canceled contracts across the country, the LAPD had suspended use of Flock cameras, and the company’s public-safety story was being re-read through immigration, privacy, accuracy, and governance concerns.[1][2]

That is the useful lesson in Flock Safety’s AI marketing crisis. Not that the company marketed badly. In many places, it marketed unusually well. It built a public-safety category frame, turned customer proof into a content engine, published trust pages where most surveillance vendors stay vague, trained police users to defend the product in public meetings, and backed the ground game with state-level lobbying. The problem is harder: each layer clarified the product, and the clearer the product became, the more some communities objected to what it made possible.

For senior marketers working around AI, surveillance, security, compliance, or other high-scrutiny categories, Flock is not a neat cautionary tale about tone. It is a case study in the limits of persuasion when the buying committee and the affected public are not the same audience.

The Business Can Be Winning While the Legitimacy Story Is Losing

The temptation is to treat contract cancellations as proof that the market has rejected Flock. That overstates what the public record supports. Reuters reported in March 2025 that Flock had reached about $300 million in annual recurring revenue and a $7.5 billion valuation, while Forbes later reported higher figures in July 2026.[3][1] The CEO has also argued that Flock continues to add customers on a net basis.[1]

The cancellation count tells a different but narrower story. Ban Flock Cameras and Common Dreams reported 82 canceled contracts across 28 states, including 39 in the first five months of 2026.[2] That is not Flock’s internal churn data, and it should not be treated as if it were. But for a marketer, the number still matters because it shows organized local resistance turning into procurement consequences. Even if revenue keeps rising, the sales motion is being forced to carry more political, legal, and reputational weight.

That tension is familiar in enterprise AI. A company can keep growing because one audience sees urgent operational value, while another audience sees the same deployment as unacceptable infrastructure. The brand does not get to choose only the friendly interpretation once the product sits in public space.

The Sharpest Tactic Was Turning Police Users Into Public Advocates

The most revealing piece of Flock’s marketing system was not a landing page. It was a webinar.

In May 2026, The Oaklandside reported that Flock had trained roughly 1,000 law enforcement officers in a single webinar on how to speak at city council meetings, including “clear ways to reframe common concerns.”[4] As a piece of category-defense work, that is tactically elegant. It identifies the moment where deals are vulnerable, finds the messengers with the most local credibility among public-safety buyers, and arms them before opponents define the meeting.

Most B2B companies call this customer advocacy. In public-sector surveillance, it becomes something more volatile. The customer is not merely a reference account; the customer is also an arm of government asking elected officials to keep or expand a technology that residents may experience as monitoring. That changes the emotional physics of the tactic.

Inside the company’s growth model, the logic is obvious. Police departments can speak concretely about stolen vehicles, shootings, missing persons, and investigative speed. They are better validators than vendor employees because they carry field authority. They can also answer council questions in the language of local operations rather than national politics. For a marketer trying to protect municipal deals, that is the right instinct: move proof closer to the decision.

But the same move can confirm the suspicion it is designed to neutralize. If residents believe surveillance infrastructure is expanding through coordinated police-and-vendor pressure, then a trained officer does not read as a neutral practitioner. The officer reads as part of the sales channel. The phrase “reframe common concerns” is especially delicate because it suggests the goal is not merely to answer objections, but to manage the room’s interpretation of them.[4]

That distinction is where many AI companies will misread the lesson. Advocacy is not wrong by default. In a complex category, buyers should hear from people who have used the product. The risk begins when advocacy substitutes for consent. A city council hearing is not a late-stage sales call. It is a governance venue where the people affected by the system may not have been part of the buying process at all.

Transparency Became an Owned-Media Surface

Flock’s trust materials deserve more credit than critics sometimes give them. The company’s “Common Concerns Answered” page addresses topics such as privacy, data retention, surveillance, and access controls in a myths-versus-facts format.[5] The page is not a buried compliance PDF. It is packaged, navigable, and written for the debate the company knows it is having.

That matters because most companies wait too long to operationalize trust content. They publish a security page for procurement, a privacy policy for counsel, and a crisis statement after the story has already hardened. Flock instead turned recurring objections into a standing persuasion asset. Privacy expert Brian Hofer told The Oaklandside that Flock and ShotSpotter were the only surveillance vendors he knew of maintaining such public-facing pages.[4]

The page also shows the promise and limit of transparency-as-shield. Audit logs, retention rules, and access controls answer a real class of objections: Who can look? How long is data kept? What happens when the system is queried? Those answers reduce uncertainty for procurement teams and public officials who need to defend governance procedures.

They do not settle the more fundamental objection: whether the public accepts the creation of a distributed license-plate surveillance network in the first place. A clean audit trail can show that access happened under a documented process. It cannot make every community comfortable with the existence of the database, the network effects of camera sharing, or the possibility that today’s permitted use becomes tomorrow’s contested use.

This is the trap for AI trust marketing. The company thinks it is answering the concern because the FAQ maps to the objection’s surface language. The audience may be making a deeper judgment about institutional power, downstream use, or the moral acceptability of the system. In that setting, a polished trust page can help the company look more prepared while leaving the central conflict intact.

Case Studies Answered the Use-Value Question, Not the Social-License Question

Flock’s “Solved Stories” and “Solved Stories Awards” are the most conventional part of the playbook. They do what good B2B case-study engines do: attach the product to emotionally legible outcomes, elevate customer operators, and make value visible through incidents rather than feature lists.[6]

For police buyers, that content has obvious force. A story about a recovered vehicle or a closed investigation does not require an abstract ROI model. It says the technology helped someone act faster. In public safety, speed and specificity are powerful conversion assets.

But these stories operate on a narrower question than the public controversy. They can support the claim that the product helps law enforcement solve certain cases. They do not prove that the surveillance architecture is democratically acceptable, that sharing rules will remain narrow, or that people who never opted into the system are comfortable being scanned as part of routine movement.

This is not a weakness of case studies in general. It is a mismatch between proof format and objection type. A use case answers, “Does it work for the buyer?” The opposition is often asking, “What does working at scale do to everyone else?”

Local Reporting Reframed the Product Faster Than the Brand Could

The Guardian’s April 2026 investigation showed how the same cameras were being interpreted differently across cities as residents, officials, and privacy advocates questioned whether Flock’s network had outgrown local oversight.[7] That is where a national brand frame starts to fray. “Public safety” is broad enough to open doors, but local controversy is specific: which cameras, which agencies, which data-sharing rules, which neighboring jurisdictions, which past abuses, which residents already feel over-policed.

NPR’s February 2026 reporting added the immigration-surveillance layer, describing concerns that Flock data could be used in ways communities had not understood or approved when contracts were signed.[8] That concern is especially damaging to the standard vendor narrative because it shifts the debate from crime-solving to secondary access. The marketing claim may center on local police investigations. The public concern may center on who else can benefit from the infrastructure once it exists.

Once that frame takes hold, the company’s strongest assets become harder to use. A police success story may make one audience more confident while making another more anxious about the scope of surveillance. A trust page may reassure a city manager while prompting an activist to test whether the written safeguards match real data flows. A trained police advocate may help pass one vote while creating a clip that opponents use in the next town.

This is not simply a communications failure. It is a distribution problem for meaning. The company can distribute its preferred story, but communities, journalists, activists, and affected residents distribute competing interpretations that are grounded in their own risks.

Accuracy Failures Punctured the Cleanest Version of the Promise

For any AI surveillance product, accuracy is not just a product metric. It is a moral claim embedded in the sales pitch. If the system is positioned as a force multiplier for law enforcement, then false positives are not a minor user-experience defect. They become encounters, stops, penalties, or explanations demanded from people who did not choose to participate.

Business Insider’s March 2026 investigation reported examples of automated license plate reader misreads that led innocent drivers to be wrongly penalized.[9] The point is not that the reporting establishes a universal failure rate; it does not. The point is that concrete misread cases damage the simplest version of the product promise: that automated recognition produces cleaner, faster enforcement with manageable downside.

Marketers often treat accuracy concerns as a technical objection to be answered with model performance, human review, or workflow design. Those answers matter. But when the consequence lands on an innocent person, the communications burden changes. The issue is no longer whether the system is generally useful; it is whether the company and its customers have made the cost of error visible enough before asking the public to accept deployment.

Lobbying Can Protect the Channel While Raising the Stakes

The state-level lobbying layer belongs in the same system as the case studies, trust pages, and police advocacy. It is not separate from marketing; for a regulated or regulation-adjacent AI company, policy work is part of market creation. It shapes what can be bought, how procurement is justified, and which objections become legally decisive.

The strategic value is clear. If local fights are slowing adoption, a company has incentives to standardize rules, preempt hostile restrictions, or educate lawmakers before activists define the category. But lobbying also confirms that the product is not merely a tool purchased by individual departments. It is infrastructure whose rules may be contested above the local level.

That shift raises the communications standard. Once the debate moves from vendor selection to governance, “our customers love it” becomes an incomplete answer. The public wants to know who sets limits, who audits use, who has standing to object, and who bears consequences when the system is wrong or repurposed.

The Apology Showed the Brand’s Internal Pressure

The July 2026 CEO apology mattered because it collapsed the distance between corporate messaging and emotional posture. Forbes reported that Flock CEO Garrett Langley had called DeFlock a “terroristic organization” in September 2025 and apologized in July 2026.[1] In an ordinary vendor dispute, harsh language might pass as founder defensiveness. In a surveillance fight, it lands differently because the critics are not just competitors or online hecklers. They are part of the public whose consent the company needs, directly or indirectly.

That is why the apology belongs in a marketing analysis rather than a separate executive-communications file. It revealed how quickly a company defending safety can appear to delegitimize dissent. For a brand already facing privacy and immigration-surveillance concerns, that posture gives opponents a simpler story: the company does not merely disagree with critics; it treats them as threats.

No trust page can fully offset that signal. Governance language asks the public to believe that power will be used carefully. Public contempt for watchdogs asks the public to believe something else.

The Lesson for AI Marketers Is Not “Be More Transparent”

Flock Safety’s marketing system did many things growth-stage AI companies are told to do. It made the value proposition concrete. It armed customer champions. It answered objections in owned media. It collected stories. It treated policy as part of market development. None of those moves are inherently cynical. In a less contested category, the same machinery would be admired as disciplined category building.

The harder lesson is that trust work increases exposure to the product’s core premise. If the audience’s real concern is confusion, better explanation helps. If the audience’s real concern is the product becoming more understandable, more connected, and more durable, better explanation can sharpen the objection.

That distinction should sit upstream of the campaign plan. Before scaling advocacy, case studies, lobbying, and trust content, AI companies need to ask whether the thing being marketed can survive the audience understanding it more clearly. Sophisticated marketing can clarify value, mobilize allies, and buy time. It cannot permanently resolve a mismatch between the product’s central promise and the public’s changing definition of acceptable risk.

References

1. Flock Safety CEO Apologizes For Calling Activists “Terrorists”, Forbes, July 2026.
2. Campaigners Celebrate as 82 Flock Camera Contracts Canceled Across 28 States, Common Dreams, 2026.
3. Flock Safety Valuation and ARR Report, Reuters, March 2025.
4. Flock Safety trained police to speak at city council meetings, The Oaklandside, May 2026.
5. Common Concerns Answered, Flock Safety.
6. Solved Stories Awards, Flock Safety.
7. Why cities are shutting down Flock cameras amid privacy concerns, The Guardian, April 2026.
8. Immigration surveillance concerns drive Flock camera contract cancellations, NPR, February 2026.
9. ALPR misreads wrongly penalized innocent drivers, Business Insider, March 2026.