An automated licence plate reader combines a high-speed camera with image-processing technology to capture licence plate numbers and enable surveillance by the collection and retention of bulk data. The data collected includes date, time and location, and may also include images of the vehicle, the driver and any passengers. They are commonly attached to police vehicles and mobile trailers, or mounted on infrastructure such as street lights and highway overpasses. They are also used by local governments for electronic toll collection on bridges and by highway authorities to catalogue the movement of traffic. The use of this technology has been coming under increasing scrutiny although the jurisprudence in Canada and the United States has yet to fully explore the privacy implications of this kind of data collection and retention.

1. Introduction

On September 1, 2020, the Citizen Lab, an interdisciplinary laboratory at the Munk School of Global Affairs & Policy, and the International Human Rights Program, both at the University of Toronto, released a report titled To Surveil and Predict: A Human Rights Analysis of Algorithmic Policing in Canada. The report examined two broad categories of algorithmic policing technology: (1) predictive policing technology (including both location-focused and person-focused algorithmic policing); and, (2) algorithmic surveillance technology. (here) The types of algorithmic surveillance technologies examined in the report included automated licence plate readers (ALPRs) that do not inherently have a predictive element. But, as Miles Kenyon said in a summary of the report titled Algorithmic Policing in Canada Explained, ALPRs provide police agencies with “sophisticated generalized surveillance and monitoring functions.” (here)

The technology is highly sophisticated. ALPRs do more than collect licence plate numbers. They enable surveillance. As Mr. Kenyon stated, ALPRs enable surveillance by “systematically collecting data in bulk, such as the date, time, and geolocation of all scanned vehicles, and the registration information (i.e., drivers’ identities) associated with those vehicles.” The technology is in widespread use by law enforcement across Ontario, British Columbia, Alberta, Saskatchewan, Quebec, Nova Scotia and Prince Edward Island. In 2017, the Information and Privacy Commissioner of Ontario published non-binding guidelines that proposed oversight of how the devices should be used. (here)

In a recent post to CNET News, senior reporter Alfred Ng described the Total Analytics Law Officers Network (TALON), announced by Flock Safety, that has ALPR cameras in 700 cities in the United States “essentially creating a nationwide camera network for tracking car movements if they’re all connected.” The TALON network,” Mr. Ng reported, “looks to connect the 400 law enforcement agencies using its cameras, allowing agencies that opt in to view camera data from other regions.” (here) In an earlier post to the Electronic Frontier Foundation’s blog titled Courts Issue Rulings in Two Cases Challenging Law Enforcement Searches of License Plate Databases, surveillance litigation director Jennifer Lynch said that “[s]ome models can photograph up to 1,800 license plates every minute”. (here) The location data is so precise it can place a vehicle in front of a specific house or within a specific lane on a bridge. “Licence plate readers,” Mr. Ng said, “are like facial recognition for cars.”

2. Privacy Implications

Police agencies and private corporations could conceivably track a particular vehicle continuously, generating a vast amount of core biographical information about a person’s private life – where and when they travel, and how often, the locations visited and associations linked with other vehicles at the same location at the same time. Trips to the psychiatrist, an abortion clinic, a by-the-hour motel, a defence lawyer’s office, a day care centre, a church, a synagogue, a mosque – all are captured and catalogued for future retrieval and potential sharing with thousands of other agencies. The aggregation of ALPR data can paint a continuous portrait of a driver’s life for posterity.

These concerns were highlighted in 2017 by Christopher Parsons, a senior research associate at the Citizen Lab, in a guest post to the blog maintained by the Privacy Commissioner of Canada. “When you drive down the road or park your car, have you considered who might be recording where your car was at any given time, and where that information is stored and shared?” he asked. (here) ALPR systems deployed by police agencies can include databases of plate numbers of interest using a watch list. The use of a watch list will generate hits (a plate number on the list) and non-hits (a plate number not on the list). In both types of hits, the location and time is stored in a local database and sometimes uploaded to a central server. In some cases, non-hits are purged before uploading but some ALPR systems “are configured to retain even non-hit data for indeterminate periods”.

Mr. Parsons identified two critical problems with the deployment of ALPRs by law enforcement – over-collection and inappropriate retention. The impact, then, is that Canadian police agencies are collecting vast troves of private information about the public, the majority of which is not suspected of having committed any crime, just by going for a drive. The deployment of ALPRs frequently lacks particularized suspicion and the unregulated use of the technology therefore violates the right of all Canadians to be protected against unreasonable search or seizure under s. 8 of the Charter of Rights and Freedoms. The right to privacy in public spaces, often called the right to anonymity, is well established in Canadian constitutional law and was affirmed by the Supreme Court of Canada in R. v. Spencer and, more recently in R v. Jarvis, that I discussed in previous posts to On The Wire. (here and here)

3. The Developing Case Law

The deployment of ALPRs has emerged in four reported Canadian cases, three involving law enforcement and one involving a private corporation. The police agencies were the RCMP in Winnipeg (R. v. Yan, 2015 MBQB 44); the Ontario Provincial Police at Sioux Lookout near Dryden (R. v. Wesley, [2016] O.J. No. 4344); and, the Canada Border Services Agency at the port of entry in Massena, New York, and Cornwall, Ontario (R. v. Cook, 2019 ONSC 515). The private corporation was a housing complex in Toronto (R. v. Abshir, 2020 ONCJ 165). The cases involved charges of “driving over .08”, “failing to stop” under a highway traffic statute, evading customs and attempted murder. It appears that a challenge to the search and seizure by the ALPRs, in the three cases involving agents of the state, was not brought by any of the defendants.

Litigation in the United States has been more engaged with the constitutional implications. In United States v. Yang, the United States Court of Appeals, 9th Circuit, in San Francisco, recently considered a challenge to the warrantless search of an ALPR database under the Fourth Amendment. The defendant’s suppression motion was based on an extension of the ruling of the United States Supreme Court in Carpenter v. United States, 138 S. Ct. 2206 (2018) involving the warrantless search of cell phone location data that I reviewed in a previous post to On The Wire. (here) However, the panel declined to consider the Fourth Amendment issue because Yang was driving a rental car with an expired rental agreement when the search took place. He did not have a reasonable expectation of privacy in the historical location data of the rental vehicle because there was no policy or practice of the rental company permitting lessees to keep cars beyond the rental period and simply charge for the extra time. Yang therefore lacked standing to challenge the warrantless search. (See also: Commonwealth v. McCarthy, Mass. S.J.C., Case No. SJC-12750, October 2, 2019)

In the post to the Electronic Frontier Foundation’s blog that I referred to previously, Ms. Lynch argued that the conclusion of the Yang court on standing is “hard to square” with the recent opinion of the United States Supreme Court in Byrd v. United States, 138 S. Ct. 1518 (2018). The court in Byrd held that an unauthorized driver of a rental car under the terms of the agreement had standing to challenge the search of the vehicle. Nevertheless, the panel in Yang recognized that ALPRs may in time present the same privacy issues highlighted by the Supreme Court in Carpenter and was clearly concerned about the “voluminous databases” of location data that ALPRs generate. “If enough data is collected and aggregated,” the panel said at pp. 25-6, “this could have the ability to identify quickly and easily the precise whereabouts and lifestyle habits of those whose vehicle information is recorded.”

4. Conclusion

In a report titled You Are Being Tracked: How License Plate Readers Are Being Used To Record Americans’ Movements, the American Civil Liberties Union made recommendations to ensure that ALPRs are used by law enforcement for legitimate purposes without infringing on Americans’ privacy rights. (here) The ALCU called for the adoption of legislation based on seven guiding principles. Here are three principles that address the privacy implications I have discussed in this post:

• License plate readers may be used by law enforcement only to investigate hits and in other circumstances in which agents reasonably believe that the plate data are relevant to an ongoing investigation. The police must have reasonable suspicion that a crime has occurred before examining collected license plate reader data; they must not examine license plate reader data in order to generate reasonable suspicion.
• Law enforcement agencies must not store data about innocent people for any lengthy period. Unless plate data has been flagged, retention periods should be measured in days or weeks, not months, and certainly not years.
• It is legitimate to flag plate data (1) whenever a plate generates a hit that is confirmed by an agent and is being investigated, (2) in other circumstances in which law enforcement agents reasonably believe that the plate data are relevant to a specific criminal investigation or adjudication, (3) when preservation is requested by the registered vehicle owner, or (4) when preservation is requested for criminal defense purposes.

The ACLU’s principles for legislation were published in 2013, seven years ago. They are as sound now as they were then.