The unidentified speaker appeared Wednesday, August 19, during a meeting of the San Diego City Council's Public Safety and Livable Neighborhoods Committee, where opposition to the city's automated license plate reader program has intensified amid a remarkable national backlash against surveillance technology company Flock Safety.
Speaking through the unmistakable mechanical breathing of the Star Wars villain, the man sarcastically proclaimed that “the Emperor is a fan of Flock” and urged San Diego to continue using the company's technology to follow “rebel scum” as they traveled between playgrounds, pools and gymnasiums. He then said he personally needed the system to stalk an ex-girlfriend. The remarks were deliberately absurd. The controversies they referenced are not. Check out the full video at SanDiegoVille's Instagram post below.
A Washington Post investigation published this month identified at least 50 law enforcement officers accused, charged or convicted of misusing automated license plate reader systems, including officers who allegedly used them to monitor wives, girlfriends, former partners and other women. Flock's system was involved in 46 of the cases examined by the newspaper.
One former Georgia police chief allegedly searched the license plates belonging to his former girlfriend and her teenage daughter approximately 600 times. A Wisconsin officer allegedly searched vehicles associated with an ex-girlfriend and her partner 179 times. In another recently disclosed case, a Florida officer allegedly used Flock cameras more than 700 times to monitor his estranged wife.
Those cases have transformed what was once largely an abstract debate about privacy into a much more concrete question about the architecture of modern policing: what happens when a government employee sitting behind a computer can reconstruct another person's movements without physically following them, obtaining a warrant or necessarily notifying anyone that the search occurred? For San Diego, that question is particularly acute.
The San Diego Police Department operates roughly 500 automated license plate readers throughout the city, primarily mounted alongside streetlight cameras. The devices continuously capture information from passing vehicles, recording license plates together with the time and location at which vehicles are observed.
Unlike a conventional police surveillance operation, the cameras do not begin recording after investigators identify a suspect. They collect information indiscriminately.
A vehicle belonging to someone suspected of murder and a vehicle belonging to someone driving a child to school enter the same system. The investigative distinction occurs later, when an officer searches the accumulated database. That inversion is central to the civil-liberties debate surrounding automated license plate readers.
The technology effectively permits government to collect enormous quantities of location information about innocent people on the premise that some fraction of that information may subsequently become useful to law enforcement.
The ACLU of San Diego & Imperial Counties intensified its opposition immediately before Wednesday's hearing, launching a “Get the Flock Out” campaign demanding that Mayor Todd Gloria and the City Council terminate San Diego's ALPR contracts. The organization argues that aggregated vehicle-location records can reveal far more than where a car happened to be on a particular afternoon.
Over time, they can expose patterns involving where someone lives and works, where their children attend school, which doctors they visit, where they worship, whom they associate with and whether they attend demonstrations or political gatherings. That distinction matters because location data becomes more revealing as it accumulates. A single observation may disclose almost nothing. Hundreds of observations can begin to approximate a biography.
San Diego's cameras are also components of a technological ecosystem whose scale has grown with remarkable speed. Founded in Atlanta in 2017, Flock has deployed more than 120,000 AI-enabled license plate cameras nationwide and says its technology is used by approximately 7,000 law enforcement agencies. The cameras collectively generate an estimated 20 billion license plate scans each month.
The company says the system assists law enforcement in approximately one million criminal investigations annually and has emphasized cases involving abducted children, stolen vehicles, violent offenders and missing people. There is little dispute that automated license plate readers can be extraordinarily effective investigative instruments.
San Diego police have credited their system with more than 400 arrests and approximately $6 million in recovered stolen property, while vehicle theft reportedly declined approximately 20% between 2023 and 2024.
The public-policy dilemma is therefore considerably more complicated than a binary argument over whether the technology “works.” Its effectiveness is precisely what makes the question of oversight so consequential.
A database capable of rapidly locating a kidnapping suspect is powerful because it can rapidly locate people. The same characteristic that makes the system valuable during an emergency makes misuse extraordinarily invasive. And unlike traditional police surveillance, the geographic reach of an interconnected ALPR network can extend far beyond the jurisdiction in which a camera is physically installed.
Until recently, defenders of ALPR networks could plausibly characterize fears of individual abuse as speculative. That position has become substantially more difficult to maintain.
The Washington Post's examination of thousands of pages of police and court records documented cases in which law enforcement officers allegedly transformed systems designed for criminal investigations into mechanisms for monitoring intimate partners and personal acquaintances. The alleged conduct was not limited to occasional curiosity searches.
Investigators found instances of repeated, prolonged monitoring capable of revealing the rhythms of another person's daily life. Some victims reportedly became convinced they were being physically followed or electronically tracked because former partners appeared to know where they had been.
The implications are especially serious in domestic-abuse situations because the individual allegedly misusing the surveillance system may possess both government authority and privileged access to investigative tools unavailable to the person being monitored.
The controversy has now become sufficiently severe that Flock itself has acknowledged deficiencies in its safeguards. Following the Post investigation and other disclosures, CEO Garrett Langley announced that the company would make several previously optional protections mandatory. Flock plans to require criminal case identifiers for searches, automatically analyze queries for abnormal behavior and reduce its default data-retention period from 30 days to seven. The company's automated auditing system is designed to flag suspicious patterns such as repeated searches of the same vehicle or unusual bursts of queries.
The reforms represent a significant change in posture for a company that previously afforded individual police agencies considerable discretion over auditing and access controls. They also raise an unavoidable question: if those protections are now considered necessary, why were they optional while an immense nationwide vehicle-location database was being assembled?
Locally, scrutiny has expanded beyond Flock's cameras themselves. Late last year, SDPD entered a one-year pilot agreement involving Flock's Nova investigative platform without first bringing the technology through the public review process that critics contend is required under San Diego's TRUST surveillance ordinance.
The arrangement was not initially disclosed to the City Council, Privacy Advisory Board or public. After its existence was reported, SDPD maintained that Nova had not been activated and said it did not intend to use the platform. The department reportedly reasoned that Nova did not require separate surveillance review because it merely processed information generated by technologies that had already been approved.
Critics see a much more consequential distinction. According to an internal Flock presentation cited in reporting supplied to SanDiegoVille, Nova can allow investigators to move from a license plate to an individual and then establish relationships with other people, including associations based upon marriage or alleged gang affiliation. That represents an important evolution in surveillance capability.
Reading a license plate answers a relatively narrow question: Which vehicle passed this camera? Combining location information with identity records, police files and relationship data permits much broader questions: Who is this person? Where have they traveled? Who are they connected to? What patterns can be inferred from those relationships? And Flock appears to be moving rapidly toward precisely that model.
Reporting published this week by WIRED revealed details about another developing Flock platform known as OS Investigate, an artificial-intelligence system designed to combine information from vehicle cameras with police reports, dispatch records, commercial databases and identity information. According to WIRED's examination of the system's code, the platform contains dozens of predetermined investigative prompts and can potentially identify individuals based upon behavioral patterns, locations, physical descriptions and associations even when investigators do not begin with a person's name or license plate.
That development complicates one of the most persistent defenses of ALPR technology: that cameras track vehicles rather than people. At sufficient scale, and when combined with other databases, the distinction becomes increasingly tenuous.
A license plate can lead to a registered owner. An owner can lead to an address. Repeated camera observations can establish travel patterns. Those patterns can reveal recurring destinations. Other databases can establish relationships. Artificial intelligence can then search across those datasets at a speed and scale that would have been impossible for human investigators.
The result is not merely a better license plate reader. It is the foundation of an integrated investigative intelligence system. Flock is simultaneously expanding into drones, audio detection and other police technologies, signaling ambitions considerably broader than its original roadside-camera business.
Another persistent concern involves whether locally collected information can ultimately reach federal immigration authorities. San Diego officials have repeatedly emphasized California restrictions on sharing ALPR information with out-of-state and federal entities and have adopted additional limitations on access.
But controversies elsewhere have demonstrated that technological networks do not always remain confined within the boundaries contemplated by local policymakers. Illinois regulators determined last year that Flock violated state law through a pilot program that provided U.S. Customs and Border Protection access to camera information in the state. Civil-liberties organizations have also raised concerns about the potential use of vehicle-location databases to identify people traveling for reproductive health care, immigration-related purposes or political activity.
The issue is not simply whether SDPD currently intends to conduct such searches. Surveillance infrastructure generally outlives the political circumstances under which it was authorized. Mayors change. Police chiefs retire. Prosecutorial priorities shift. Federal administrations adopt different enforcement policies. Laws are amended. Vendors introduce new features.
Data collected for one purpose can acquire entirely different significance under another government. That temporal dimension is often absent from debates about surveillance technology. Policymakers are asked whether they trust the officials operating a system today when the more durable question may be whether they are comfortable constructing a system that officials will inherit tomorrow.
The potential harm is also not confined to deliberate misconduct. San Diego resident Hugo Parra spent approximately 28 days incarcerated after being arrested in connection with an attempted armed carjacking following a Flock alert and witness identification. According to reporting about Parra's pending litigation, however, Flock's own location records placed his vehicle approximately five miles from the crime scene when the offense occurred.
“The Flock alert clearly indicated the wrong vehicle - it was impossible for it to be in both locations at once,” Parra's attorney, Alex Coolman, told Ars Technica.
The case illustrates a different vulnerability inherent in algorithmically assisted policing. Technology does not need to malfunction spectacularly to produce injustice. A camera can accurately photograph a vehicle. Software can correctly read a plate. A database can faithfully report its location.
The failure can occur afterward, when human beings assign excessive evidentiary weight to an automated lead, interpret ambiguous information as confirmation or allow technological authority to reinforce an existing suspicion. Twenty-eight days of incarceration is an enormous consequence for that kind of error.
San Diego's relationship with Flock has also attracted attention because of the movement of a former senior police official into employment with the company. Former SDPD Captain Jeff Jordon helped oversee the department's adoption of advanced surveillance technology before retiring and subsequently taking a position with Flock. There is no public evidence that Jordon violated an ethics law, received compensation from Flock while employed by the city or improperly influenced a procurement decision.
Nevertheless, Councilmember Henry Foster III publicly criticized the optics during last year's contentious review of the surveillance program, calling the situation “totally inappropriate.” The controversy reflects a broader structural issue in the rapidly expanding police-technology industry.
Former law enforcement officials possess exactly the expertise surveillance companies need: familiarity with departmental procurement, operational requirements, government budgeting and the institutional arguments most likely to persuade police agencies and elected officials. Their movement into private-sector positions is not inherently improper. But when officials involved in evaluating or implementing a technology later work for the company selling it, public confidence depends heavily upon transparent procurement procedures and meaningful independent oversight.
Perhaps no city should understand that problem better than San Diego. The city's original smart-streetlight system was initially presented principally as municipal infrastructure capable of collecting information useful for traffic planning and other civic purposes. Police access to those cameras eventually generated an intense controversy after the public learned how extensively the technology could be used for law enforcement.
A city audit subsequently identified serious deficiencies in oversight and governance. The fallout helped produce San Diego's TRUST ordinance and the Privacy Advisory Board — mechanisms intended specifically to ensure that surveillance technology would not again become operationally embedded before the public understood its capabilities. That history makes the dispute surrounding Nova especially significant.
The central question is no longer simply whether San Diego has written rules governing surveillance. It is whether those rules can keep pace with technology whose capabilities may change through software rather than the installation of new physical equipment.
A camera approved to read license plates can become substantially more powerful when its information is connected to another database. A database can become more powerful when artificial intelligence can interrogate it. An investigative system can become more intrusive when another jurisdiction gains access. The hardware may remain unchanged while the surveillance capability becomes almost unrecognizable.
The backlash is no longer confined to traditional civil-liberties organizations or politically progressive cities. Communities around the country have begun terminating, rejecting or reconsidering Flock contracts. More than 50 jurisdictions have reportedly severed relationships with the company this year, while opposition has emerged from groups spanning the ideological spectrum.
The potential harm is also not confined to deliberate misconduct. San Diego resident Hugo Parra spent approximately 28 days incarcerated after being arrested in connection with an attempted armed carjacking following a Flock alert and witness identification. According to reporting about Parra's pending litigation, however, Flock's own location records placed his vehicle approximately five miles from the crime scene when the offense occurred.
“The Flock alert clearly indicated the wrong vehicle - it was impossible for it to be in both locations at once,” Parra's attorney, Alex Coolman, told Ars Technica.
The case illustrates a different vulnerability inherent in algorithmically assisted policing. Technology does not need to malfunction spectacularly to produce injustice. A camera can accurately photograph a vehicle. Software can correctly read a plate. A database can faithfully report its location.
The failure can occur afterward, when human beings assign excessive evidentiary weight to an automated lead, interpret ambiguous information as confirmation or allow technological authority to reinforce an existing suspicion. Twenty-eight days of incarceration is an enormous consequence for that kind of error.
San Diego's relationship with Flock has also attracted attention because of the movement of a former senior police official into employment with the company. Former SDPD Captain Jeff Jordon helped oversee the department's adoption of advanced surveillance technology before retiring and subsequently taking a position with Flock. There is no public evidence that Jordon violated an ethics law, received compensation from Flock while employed by the city or improperly influenced a procurement decision.
Nevertheless, Councilmember Henry Foster III publicly criticized the optics during last year's contentious review of the surveillance program, calling the situation “totally inappropriate.” The controversy reflects a broader structural issue in the rapidly expanding police-technology industry.
Former law enforcement officials possess exactly the expertise surveillance companies need: familiarity with departmental procurement, operational requirements, government budgeting and the institutional arguments most likely to persuade police agencies and elected officials. Their movement into private-sector positions is not inherently improper. But when officials involved in evaluating or implementing a technology later work for the company selling it, public confidence depends heavily upon transparent procurement procedures and meaningful independent oversight.
Perhaps no city should understand that problem better than San Diego. The city's original smart-streetlight system was initially presented principally as municipal infrastructure capable of collecting information useful for traffic planning and other civic purposes. Police access to those cameras eventually generated an intense controversy after the public learned how extensively the technology could be used for law enforcement.
A city audit subsequently identified serious deficiencies in oversight and governance. The fallout helped produce San Diego's TRUST ordinance and the Privacy Advisory Board — mechanisms intended specifically to ensure that surveillance technology would not again become operationally embedded before the public understood its capabilities. That history makes the dispute surrounding Nova especially significant.
The central question is no longer simply whether San Diego has written rules governing surveillance. It is whether those rules can keep pace with technology whose capabilities may change through software rather than the installation of new physical equipment.
A camera approved to read license plates can become substantially more powerful when its information is connected to another database. A database can become more powerful when artificial intelligence can interrogate it. An investigative system can become more intrusive when another jurisdiction gains access. The hardware may remain unchanged while the surveillance capability becomes almost unrecognizable.
The backlash is no longer confined to traditional civil-liberties organizations or politically progressive cities. Communities around the country have begun terminating, rejecting or reconsidering Flock contracts. More than 50 jurisdictions have reportedly severed relationships with the company this year, while opposition has emerged from groups spanning the ideological spectrum.
The unusual coalition reflects the breadth of the underlying concern. Progressive organizations have emphasized immigration enforcement, reproductive freedom, racial disparities and political protest. Libertarians and conservatives have raised Fourth Amendment concerns, government overreach and the prospect of centralized tracking. Their political reasoning differs.
Their underlying fear is remarkably similar: infrastructure capable of documenting the movements of millions of people should not become ubiquitous merely because each individual camera appears innocuous. That debate extends beyond Flock. Cities that eliminate the company's cameras increasingly replace them with competing ALPR systems, prompting privacy advocates to argue that focusing exclusively on a single controversial vendor risks obscuring the larger question of whether pervasive vehicle surveillance should exist at all.
Flock's strongest argument is also the reason scrutiny of the company is unlikely to disappear. The technology can work extraordinarily well.
A system that instantly identifies a stolen vehicle, locates an abducted child or helps investigators find a violent offender offers tangible benefits that are easy to understand and difficult for elected officials to relinquish. But efficacy alone has never been the appropriate measure of government surveillance.
Many intrusive investigative techniques would make policing easier. Democratic societies impose limits because the government's ability to investigate crime exists alongside constitutional protections against unreasonable searches, freedom of association and the presumption that ordinary citizens should be able to move through society without continuously generating government-searchable records of their whereabouts.
Flock now occupies the center of that unresolved conflict. Its cameras have become cheaper, more numerous and more interconnected while its software has become substantially more sophisticated. The company is simultaneously responding to documented abuses with stronger safeguards and developing artificial-intelligence systems capable of extracting considerably more intelligence from the information its network collects.
San Diego therefore faces a question much larger than whether several hundred cameras help police solve crimes. It must decide what kind of surveillance infrastructure it is willing to normalize, what limits should exist before that infrastructure becomes indispensable, and whether safeguards imposed today can meaningfully restrain capabilities that may be radically different tomorrow.
Against that backdrop, Wednesday's Darth Vader appearance was less an eccentric interruption than an unusually theatrical manifestation of a serious civic dispute. The costume supplied the metaphor: an all-seeing government pursuing its enemies across an empire.
But stripped of the science-fiction references, the underlying warning was straightforward. A surveillance system does not need to be designed for authoritarian purposes to create authoritarian capabilities. It merely needs to become sufficiently comprehensive, sufficiently interconnected and sufficiently difficult to dismantle.
San Diego is now being asked to determine where that line lies, after much of the infrastructure has already been installed.
Originally published on August 20, 2026.
