Local law enforcement agencies’ growing reliance on artificial intelligence-based data reduction and presentation programs is alarming civil liberties advocacy groups.
In recent weeks and months, the Redlands and Rialto police departments have renewed and intensified their existing contracts with Axon Enterprise Inc. to replace and expand their body-worn camera programs, digitized evidence storage, logging, retrieval, sharing and management systems, add new digital evidence capabilities and outfit them with drone surveillance equipment and operational technologies and virtual reality applications.
At present, the Barstow, Chino, Colton, Fontana, Montclair, Ontario, Redlands, Rialto and, San Bernardino, Upland police departments have Chino contracts with Axon as does the San Bernardino County Sheriff’s Department.
Both the San Bernardino Sheriff’s Department and the San Bernardino Police Department have contracts with Palantir Technologies, Inc.
Palantir Technologies and Axon Enterprise are both significant players in the technology sector in the realm of data analytics. The two companies have products or services relating to the application of data processing in the realm of public safety applications.
Shortly after Donald Trump became president, federal officials approached Palantir, which had landed its first contract with the U.S. Government in 2008 and had obtained over $1.8 billion in governmental contracts over the next sixteen-and-a-half years, about creating a “master databank” for the U.S. government to be accessible to a multitude of agencies, including the Department of Defense [since renamed the Department of War], the Justice Department, the Department of Homeland Security, the Department of Health and Human Services, the Internal Revenue Service, the Department of Immigration and Customs Enforcement. Amidst reports that Palantir was being commissioned to centralize and organize data relating to Americans’ bank accounts, student debt, medical histories, disability statuses.
Axon almost entirely initially established its relationships with the county’s local law enforcement agencies as a consequence of supplying them with bodyworn cameras.
Famously, beginning in 2013, Rialto was offered the opportunity to participate in a pilot program involving Axon body cameras. When an analysis of the results of that program, in which… dozens, then scores, followed by hundreds and now thousands of police departments around the country were sold on the concept of outfitting their officers with bodyworn cameras, which has led to Axon being very profitable. While Axon stock was selling for $8.94 a share on January 2, 2013 and $15.88 a share on December 31, 2013, one share today goes for $612.98. While the cameras are Axon’s leading product, it also offers software and service to augment the video those cameras provide to allow the footage to be put to use, primarily as evidence in a prosecutorial setting, as well as for internal departmental purposes, such as departmental higher ups, uniformed commanders and administrators monitoring the performance of the officers wearing those cameras. The company has thus evolved and/or expanded into divisions which provide data storage and retrieval capability, along with the capability of analyzing that data as well as cataloging it so it can be distinguished from the millions, indeed billions, of files accumulated on a continuous basis. In 2018, Axon moved into the provision of drone aircraft to law enforcement agencies. In 2019, Axon introduced its Axon Fleet 3 in-car video system, which featured automated license plate recognition capability. A sideline to the Fleet 3 and later license plate readers developed by the company was the data storage, analysis and retrieval system that made the information being accumulated by the cameras useful to the department’s in applying it to investigations and solving cases or apprehending suspects and getting convictions against them in a court of law.
What Axon was doing in this venue mirrored, approximated, overlapped and paralleled much of what Palantir was doing within the sphere of servicing its law enforcement and public safety agency clients.
Palantir, however, was not oriented to simply serving law enforcement or a combination of law enforcement and the military, but a whole host of both public sector and private sector customers that were not only involved in accumulating seemingly unlimited volumes of data but in need of storing the information digitally and then being able to manage it and wield it by analyzing the data and filtering from it something that is usable for in the context of whatever objective that entity – be it a government or government agency or large corporation or medium or small sized company – was pursuing.
Both Palantir and Axon over the years have developed several iterations – or generations – of data storage and data analytical programs, with each succeeding one growing more comprehensive and sophisticated, with ever greater capacities for enlargement, merging and cross-referencing of databases.
One such highly evolved program from Palantir is its Gotham software platform designed to integrate, manage, and analyze massive datasets for law enforcement agencies and enable them to uncover hidden patterns and relationships to make data-driven decisions. Another Palantir program developed for specific law enforcement application is its Laser program intended to make “predictive policing” possible by identifying chronic offenders and hot spots through the compiling of gang databases.
Axon Enterprises most advanced program in this vein is its Axon’s Draft One which utilizes formulas, algorithms and artificial intelligence applied to data to link it all together and generate police reports.
Given that they are separate companies and to some degree are in competition with one another. Palantir and Axon do not directly share data. Nevertheless, bot are involved in with multiple law enforcement agencies. In some cases, those law enforcement agencies have an exclusive relationship with just one of the companies. In other cases, a law enforcement agency may have a contract with both of the companies. In such cases where a law enforcement agency has a contractual arrangement with both companies, that agency at least with regard to some investigative matters and enforcement undertakings stands to benefit from integration of the data held by by both Palantir and Axon and the analysis of that data once it has been integrated or merged. In this way, at least within the context of such specific law enforcement agencies that contract with the two companies, the Palantir and Axon databases have been merged.
Moreover, given the common protocol of cooperation between police departments and law enforcement agencies in general – a principle referred to as agency-to-agency privilege – the merging of the respective databases that are managed by Palantir-provided or Axon-provided software or software provided by any supplier can and does take place.
Civil liberties advocacy groups, which have weighed in with regard to the threats the integration of databases and the use of algorithms and artificial intelligence police departments in New York City, Chicago, New Orleans, San Francisco, San Jose, Oakland and Los Angeles to analyze the data and draw conclusions which are then presented as reliable probable cause to proceed to the next step in the investigative or legal process represent to citizens’ constitutional rights, are now expressing similar concerns with regard to the way police departments and the sheriff’s department in San Bernardino County are similarly combining available datasets in a way that is comprehensively invasive, such that not only are the everyday comings and goings of virtually everyone living in the county’s 22 cities, two incorporated towns and 87 unincorporated communities virtually instantaneously knowable by working law enforcement officers, but, potentially, their banking records, credit card expenditures, medical records and associations, including those who are and are not involved in illegal or suspicious activity.
While the U.S. Constitution prohibits governmental entities from eclipsing the rights of American citizens to obtain information pertaining to them deemed private and confidential without agents of the government first obtaining a warrant from a court to access that information, there is nothing in the Constitution to prohibit the government from monitoring or gathering information, images or sounds in public spaces using photographic, videographic or other technical means or capturing in the open ether through which cellphone communications propagate radio signals. The government is not enjoined from purchasing from entities in the private sector data those companies have accumulated within a commercial context from other companies or service providers which are involved in providing both services and commodities to the public.
In response to articles that appeared in the New York Times last year detailing the company’s construction of a “master database” for surveillance in service of the federal government during the Trump Administration, Palantir characterized reports that it was serving as a means for government officials to access citizens’ banking, medical, interpersonal interactions, online communications, travel and telecom data as “fiction… blatantly untrue [and] reckless.” According to Palantir, in providing the government the means and technology to integrate a myriad of databases to be able to draw accurate conclusions based upon multiple points of information and contexts is a service to the general public. Palantir said its algorithims protects rather than violates while remaining in compliance with ensuring the company does not trample on civil liberties and adheres to all applicable legal standards.
According to the federal government, its contractual relationship with Palantir represents a substantial boon to U.S. citizens and will expedite the speed with which they obtain any benefits they are eligible and apply for, simplifying their tax payments, improve the fair and judicious employment of the justice system against the innocent and guilty alike and lead to the uncovering of waste, fraud and abuse in and of governmental programs.
Critics, however, point out that Palantir’s first loyalty is to its bottom line and stockholders, its second loyalty is to its customers and clients and any concerns beyond that are tertiary. Palantir’s paying customers are purchasing services that the company can provide and it would be grossly naïve to believe that Palantir would withhold information from a client that was paying hundreds of millions or tens of million or even just millions of dollars to ensure that it has access to information that Palantir has access to.
The American Civil Liberties Union in 2024 published a report outlining its organizational concerns with regard to the use of artificial intelligence in the data selection and data reduction process that goes into the production of integrating data and drafting police reports. Palantir’s Laser program renders “data visualization documents” as well as “predictive alerts” and Axon’s Draft One program uses ChatGPT to translate body camera recordings into initial report drafts. According to the American Civil Liberties Union, law enforcement agencies use these artificial intelligence driven systems to engage in predictive policing and deliver reports that are machine written rather than composed by officers.
According to the American Civil Liberties Union’s senior policy analyst Jay Stanley, artificial intelligence systems are prone to unpredictable errors and biases, which can be compounded by by poor-quality audio from body cameras. According to the American Civil Liberties Union, the use of artificial intelligence systems to bypass the time-consuming process of having officers write their own reports deprives the justice process of the actual version of events that can be provided by those officers, including particular nuances that might escape the parameters of an algorithm. The American Civil Liberties Union maintains that officers’ memories should be memorialized in a traditional report before those officers’ memories are influenced by artificial intelligence-generated narratives.
The American Civil Liberties Union also objects to the opacity of the programming of artificial intelligence-driven integration, management and analyzation of datasets acquired by law enforcement agencies and the logic that leads to the conclusions relating to investigative direction and probable cause assertions or assumptions leading to warrant issuing or further investigative action, and how those investigative workings can be scrutinized fairly with regard to constitutional issues in legal proceedings subsequent to an arrest or in a civil action following the dismissal of criminal charges following an arrest.
In addition, the American Civil Liberties Union has privacy concerns, with questions about the handling and storage of sensitive police data used to form the basis of the artificial intelligence models used or to train officers in utilizing them. Because police reports represent a central element in criminal proceedings, often serving as the primary official account of incidents, the American Civil Liberties Union has expressed misgivings that artificial intelligence-generated reports could omit important details or even provide cover for potential misconduct not captured on camera in cases growing out of reports based primarily on bodyworn camera footage.
A number of problems have surfaced with regard to Palantir’s programs, at least a partial element of which relates to the high cost of the service Palantir offers. Because of the cost, at least some departments have purchased only what amounts to a “partial” service, such that the departments “fill the gap” with non-Palantir backed conclusions or algorithms or the departments’ own makings, in which the case the reliability of this hybrid is completely unknown and virtually unchartable.
The Los Angeles Police Department utilized Palantir’s Laser program for predictive policing, seeking to identify chronic offenders and crime hot spots. The department abandoned the program in 2019 due to civil rights concerns, criticism that the system was reinforcing racial bias, and questions about its effectiveness and oversight.
The New York Police Department utilized Palantir’s Gotham system to integrate vast and separate databases such as arrests, tickets and social media postings into a searchable network, enabling complex analysis that led to legitimate arrests and breakthroughs in solving some crimes, but also provoked major concerns about privacy, racial bias from historical data, and the lack of transparency in giving a private company access to public data. Palantir’s performance in conjunction in connection with the NYPD contract was subject to criticism over the company’s alleged misuse of the data entrusted to it amidst debates about data ownership and expanding public and private surveillance, issues which have not been fully resolved, according to the Brennan Center for Justice.
The Chicago Police Department used Palantir’s Laser (Law Enforcement Analysis and Reduction) program and predictive model software to manage its gang database and create “chronic offender scores,” identifying individuals for focused police attention by analyzing criminal history, social media, and vehicle data. The department claimed that this proactively prevented violence, which was to some extent true. Nevertheless, critics assailed the department for unfairly targeting minority communities, which led to controversy and scrutiny with regard to the program’s actual effectiveness and fairness. The department’s efforts at predicting crime, those chastising the department said, were actually racial profiling. The effectiveness of the data analysis was vindicated by the number of arrests, though detractors said that no matter how well data points were connected and suspects rounded up, the operations did not result in sufficient evidence being obtained as a consequence of the arrests to get court convictions and most cases ended in reduced pleas.
The New Orleans Police Department used early versions of Palantir’s data mining service beginning in 2012, doing so secretly such that members of the city council did not know the predictive policing program, referred to as Project Nimbus, existed. Project Nimbus utilized software hashing through a variety of data such as ties to gang members, criminal histories, and social media to predict the likelihood that a select group of roughly 3,900 potential victims and perpetrators would commit acts of violence or become victims. Palantir’s prediction model consisted of evaluating social network exchanges and postings, individuals, places, vehicles, weapons, addresses, and information contained in field interview cards, references to which were previously were held in separate databases. The program, which was extended three times through February 21, 2018 corresponded to a temporary drop in violent crime, though the department and the district attorney’s office have said it’s not clear that Palantir’s program was responsible for that reduction. When the existence of the project was revealed, there was widespread outrage over what were perceived as undue secrecy and violations of privacy.
Palantir undertook a predictive policing pilot program in Santa Cruz in 2011, utilizing data to predict and reduce property theft and gang activity. After an early show of success in which officers received timely alerts relating to car theft, controversy ensued as there was scrutiny and concern with regard to alleged algorithmic bias and the perpetuation of racial profiling and inequality, ultimately leading, after a prolonged debate, to the city banning such technology.
In the course of all of this, the term “predictive policing” has become an undesirable label. As a consequence, Palantir and the agencies with which it contracts substitute terms such as “intelligence-led policing” or “data-driven analysis.”
All of this debate about potential violations of citizens’ privacy and civil rights has not prevented local agencies from entering into or intensifying their contracts with Axon and Palinter.
On November 25, 2025, the Rialto City Council approved having the city enter into a nine-year, $14.3 million augmentation contract with Axon Enterprise Inc., which is also known as TASER International, to increase an existing contract it had for supplying the police department with surveillance, video, information processing and storage, software, and artificial intelligence technology. The package the city purchased, Axon’s Officer Safety Plan 10 (OSP 10), includes making upgrades to the body-worn and in-car camera system that has long been deployed by the department, the company’s next generation of digital information storage and retrieval system, automated video analysis, license-plate readers and the Rialto Police Department’s first Drone-as-First-Responder program.
Rialto since 2012 has had a contractual relationship with Axon for the provision of bodyworn cameras used by the police department’s officers.
While the Drone-as-First-Responder fleet is to consist of unmanned surveillance craft provided through the company’s Axon Air division, the department already had in place three drones as part of the department’s unmanned aircraft system, which has been referred to by using the nomenclature “UAS/Drone.”
According to the department, it has in its inventory a single DJI Mavic 2 PRO remote-controlled aircraft, purchased at a cost of $1,500, and two EVO 11 PRO remote-controlled aircraft, purchased for a total cost of $3,750, which are “utilized to enhance the safety of the community and officers.” Each of those drones, equipped with video cameras, have come into use, according to the department, “when its view would assist officers or incident commanders with the following situations, which include but are not limited to 1) major collision investigations; 2) search for missing persons; 3) natural disaster management; 4) crime scene photography; 5) SWAT [special weapons and tactics] tactical or other public safety and life preservation missions; 6) in response to specific requests from local, state or federal fire authorities for fire response and/or prevention.”
On December 16, 2025, the Redlands City Council approved a $3,544,704 agreement with Axon to replace city police department’s existing body-worn camera and drone programs with two different vendors and include digital evidence, drone and virtual reality technologies to the follow-on arrangement it is entering into with Axon.
Included in the $3.54 million package was the next generation of body-worn camera equipment and software, a drone program, virtual reality training tools and Axon’s Justice Suite. The vote included an appropriation of $708,940 from the general fund and the police department’s asset forfeiture fund to cover a portion of the costs.
The agreement dispenses with the department’s body-worn camera system, secured in 2022 under a five-year lease with another company, LensLock, for 110 cameras. The new system, according to the department, does not have limitations in the evidence uploads of the previous cameras and accompanying software, which require that officers dock cameras at the police station to upload video. This could entail a two-week delay in getting evidence to the San Bernardino County District Attorney’s Office, officials said. The footage could not be reviewed or shared immediately, and was difficult to share both with the district attorney’s office and other entities entitled to view such footage, such as federal agencies and defense attorneys. The district attorney’s office uses Axon Justice for its digital evidence management, compiling, transcription and redaction. The department’s LensLock body-worn camera system was incompatible with that platform, requiring prosecutors to manually import data. City officials said the new agreement will allow evidence to be shared securely within hours rather than weeks and reduce the time department personnel must devote to manual processing of the data.
The Redlands Police Department justified the transition by noting it processes roughly 1,441 video evidence or discovery requests per year, and spends on average 1.25 hours of personnel time handling each request. The transition to Axon Evidence will automate that workflow and reduce overtime, the city maintains. The new body-worn cameras include a real-time translation feature, automatic dictation of recorded audio, and geo-positioning system tracking, along with automatic activation.
The Redlands Police Department elected to go all-in with Axon, canceling its previous contract with LensLock for the body worn cameras, its contract with DJI for drones and its contract with Flock/Motorola for license plate readers in favor of the comprehensive contract with Axon. Axon will use Skydio drones to supplant the DJI drones. The Skydio drones, which have live aerial and thermal imaging capability, will be stationed in enclosed docking units on city-owned rooftops. They are to be used for so-called first response to reconnoiter situations when officers are en route to incidents that could be dangerous.
“No other vendor offers a comparable integrative solution with the same level of interoperability,” said Redlands Police Department Chief Rachel Tolber. “Axon’s platform is widely adopted across law enforcement agencies and is considered the industry standard amongst localized partners and agencies.”
The agreement extends to the employment of the Axon Fusus platform, which integrates live data from body-worn cameras, drones, license plate readers and other systems into a single interface, supporting current operations and potential future real-time information center capabilities.
It was not clear how the city will settle up with Flock, Motorola, DJI and LensLock upon leaving them in the lurch.
The San Bernardino Police Department has a contract with Palantir to use its Gotham platform for the integration and analysis of various data with regard to city residents, allowing it to build intelligence profiles and identify individuals and rank their likelihood to be involved in criminal activity and whether they might be chronic or serious offenders even if they do not have criminal records, using characteristics such as tattoos, immigration status, bank account information, place of residence and whereabouts on a constant basis or within a certain timeframe.
The San Bernardino Police Department also has a $750,000 drone first responder program, funded with $562,500 allocated to the city by the County of San Bernardino and $187,500 provided to the city by a Citizens’ Option for Public Safety (COPS) grant.
The Ontario Police Department’s Real Time Information Center utilizes drones to improve police department response and increase department members’ understanding of the circumstances they are about to encounter.
-Mark Gutglueck