UPDATE—February 19, 2014:

According to the Washington Post, just days after the story broke, DHS shelved its plans to create or tap into a national database of license plate recognition data. According to an Immigration & Customs Enforcement spokeswoman, the solicitation "was posted without the awareness of ICE leadership" and "will be reviewed to ensure the path forward appropriately meets [the agency's] operational needs.”

However, DHS may still be accessing national license plate data—collected by the private company Vigilant Solutions—on an ad hoc basis. According to documents obtained by the ACLU of Massachusetts, ICE agents and other branches of DHS have been tapping into Vigilant’s data sets for years. Like the ACLU, we'll be interested to see whether DHS and ICE release more information about their plans—and we'll continue to push back on any future attempts to create a federally-accessible national license plate database.

ORIGINAL POST:

The Department of Homeland Security and its component Immigration and Customs Enforcement recently issued a solicitation for bids to build and maintain a national database of motor vehicle license plate data. Not only would this database include plate data collected by DHS — it would also include data from other law enforcement agencies and private companies.

This is the first time any federal agency has proposed a database of this size and scope, and this "National License Plate Recognition" program raises significant privacy concerns. As we’ve said before, this kind of license plate data is location data — it tells the data gatherer where you’ve been and when, and can be aggregated to present a detailed picture of your life and who you associate with—whether you’re at a lawful protest or house of worship; a gay bar or your doctor’s office; your brother’s house or your lover’s. License plate data allows the data gatherer to track all movement in and out of an area; specifically target certain neighborhoods or organizations; or place political activists on hot lists so that their movements trigger alerts.

A Massive Expansion of Plate Data Collection

Automated License Plate Reader or ALPR cameras already scan and record the plates of millions of cars across the country. Law enforcement agencies in large metropolitan areas like Los Angeles and New York have databases of millions of plates—and these databases will only increase in size over time.  A 2011 survey of more than 70 police departments showed that 79 percent used ALPR technology and 85 percent expected to acquire or increase use in the next five years. On average, these agencies expected that 25 percent of police vehicles would be equipped with license plate readers by 2016.

However, DHS doesn’t want to limit its data collection to law enforcement agencies. It also wants to include data from “asset recovery specialists” (repo companies) and “access control systems” (private security cameras in parking lots like malls). Private companies already collect data on a nationwide basis and may have more data than all law enforcement records combined. Vigilant Solutions states its database contains nearly 2 billion records, and MVTrac claimed it had records on “a large majority" of registered vehicles in the U.S. TLO,  another company, which was recently acquired by the credit reporting agency and data aggregator TransUnion, has a “massive database of one BILLION vehicle sightings” with “up to 50 million new sightings” added each month. While some states have tried to limit the power of these companies to collect data, they’ve fought back hard on First Amendment and other grounds.

Hot Lists, Wide Shots and Historical Data

There are several other reasons why Americans should be concerned about DHS’s plans. First, the agency wants to be able to create its own “hot lists” of suspect vehicles from the data. As we’ve seen from ALPR records we received from the Los Angeles Sheriff’s Department, officers are not required to define any individualized suspicion before putting a vehicle on a “hot list,” and it’s unclear how a vehicle would ever get off such a list. DHS proposes sharing its “hot lists” with other agencies. It also wants to be able to communicate with other users, “establish Lists submissions, flag license plates, and conduct searches anonymously.” If ICE agents can create hot lists, flag plates, conduct searches and discuss and share data anonymously, meaningful oversight of the program will be impossible. There will be nothing to prevent the kind of racial, ethnic and religious targeting we’ve seen through programs like the NYPD’s stop and frisk policy and surveillance of Muslim communities and ICE’s Secure Communities program.

Another concern is that the agency wants “a zoomed out image of the vehicle” in addition to a close-up photo of the plate. This will allow the agency to identify not just the vehicle, but also its occupants. Mike Katz-Lacabe, a San Leandro, California resident, learned just how revealing these photos can be after he requested his own license plate data from the cops and received a photograph that clearly showed him and his young daughters getting out of their car in their own driveway.

Finally, it appears that DHS wants to be able to access historical plate data through the database. This could severely hamper state and local attempts to place limits on the collection, retention and use of license plate data. For example, in Michigan and Massachusetts, legislators have proposed limiting the retention of license plate data to no longer than 48 hours. If DHS is able to aggregate these states’ data with data from other sources into a massive nationwide database, the retention limits that the states’ own citizens have proposed become futile.

Putting the Brakes on Plate Data

The Washington Post has quoted an ICE spokeswoman as saying that “the data would be collected and stored by the commercial enterprise, not the government.” This does not assuage concerns. As EFF and others have stated in the context of the NSA’s mass phone data collection, transferring data storage from the government to a private third party doesn’t solve the problem. Only meaningful limitations on collection, retention, access and sharing can do that. So far, DHS has not discussed implementing any specific limitations on its own system or explained how it will honor limitations governing the license plate data it collects from other sources. Without these limitations and greater transparency in how DHS plans to use and share the data, this program should not go forward.