Professionalism/Vigilant Solutions, Law Enforcement, and Privacy

Vigilant Solutions provides law enforcement with cutting-edge technology to make surveying individuals and groups in society easier than ever. The technology is developed for law enforcement, by law enforcement, and can improve safety for officers and communities. Due to the nature of their products, numerous privacy concerns arise in their implementation on the otherwise oblivious general population.

PLATESEARCH
PLATESEARCH is a license plate recognition platform created by Vigilant Solutions. The technology can be implemented in fixed or mobile locations and uses image processing to identify and locate all license plates that come into its range. This indiscriminate behavior causes concern among many law biding citizens. The data and analytics received from these LPR detections are stored into the company's cloud database, LEARN. This database allows users to create patterns, follow and track individuals creating a constant map of someones whereabouts. The information it collects can cause a potential for misuse and lead to the invasion of privacy.

FACESEARCH
FACESEARCH is Vigilant Solutions' facial recognition tool which provides preprocessing capabilities and analyzes faces. This technology was developed to work with any network enabled camera. LineUp, the software behind FACESEARCH, searches each video frame for a human face and catalogs it in their database. The database can be combined with other image databases to gather personal identifying information on an individual. Privacy concerns arise from this as it is indiscriminate and operates under the radar. Every individual that comes across this technology is recognized, potentially without their consent, and stored into the database.

Vigilant Solutions' Stance
Vigilant Solutions believes that their products follow all laws and do not invade individuals' privacy. According to Brian Shockley, Vigilant Solution's marketing professional, license plate readers (LPRs) do not gather personal information and U.S courts have upheld their usage in communities. A LPR only gathers information that is already known to the public and therefore does not invade anyone’s privacy. DMV records are only accessible to authorized personnel and therefore LPRs themselves cannot be used by individuals to track others. People argue that the stored plates should be deleted once a crime has not been detected, but Vigilant Solutions disagrees. With this technology a crime can be easily solved as it can lead to information on suspects and can help locate and apprehend them. Vigilant Solutions recognizes that there is a potential for their products to be misused, but affirm that law enforcement that implement them live by a clearly defined process and policies that prevent it from happening.

Privacy
Collier County sherriff’s deputy Peter Deffet was fired on February 28, 2014 after frivolous searches in the Driver and Vehicle Information Database (D.A.V.I.D.). The database stores drivers’ addresses, vehicle information, driving history and other personal information. Deffet searched 151 names including Hulk Hogan, John Travolta, and local NBC-2 news anchor Heather Turco, who he searched 8 times. Turco commented “I think eight times is a little excessive. There needs to be a check and balances and they need to have an understanding that this is a no-no.” Internal investigators found Deffet guilty of violating Sheriff's Office policy of using D.A.V.I.D. for personal use.

Police use of facial recognition software is unregulated by state law. Georgetown Law’s Center on Privacy & Technology studied 52 law-enforcement agencies’ use of facial recognition technology. None of the agencies require warrants or limit searches for serious crimes. Less than 10% have a publicly available use policy, and only 17% audit officers’ searches for improper use. The Michigan State Police was the only agency with functional auditing software. Law enforcement’s use of traditional biometrics including fingerprinting is limited to criminal arrests and investigations. Poorly defined facial recognition limitations have allowed the FBI to create a biometric network of 117 million Americans by running facial recognition scans on 16 states’ driver’s license photo databases. Most of these Americans are non-criminals. One in two Americans is in a law enforcement face recognition network.

Assembly
Internal emails from the Immigration and Customs Enforcement agency (ICE) revealed they used license-plate readers at a Texas gun show near the border to investigate smuggling. They recorded the plate numbers of cars at the Del Mar gun show using plate readers and compared them to cars that passed through the border. Buying guns and crossing the border are two legal activities. American Civil Liberties Union lawyer Jay Stanley said "gun-show surveillance highlights the problem with mass collection of data...because those two activities in concert fit somebody’s idea of a crime, a person becomes inherently suspicious.” Executive director of Gun Owners of America Erich Pratt said: “Information on law-abiding gun owners ends up getting recorded, stored, and registered, which is a violation of the 1986 Firearm Owners Protection Act and of the Second Amendment."

CEO of PlateSmart Technologies John Chigos called the event “an abuse of technology. I think this was a situation that shows we need to establish policies for license-plate readers, like any new technology.” Bob Templeton is the CEO Crossroads of the West, the company that organizes the gun show. He says ICE’s actions are “obviously intrusive and an activity that hasn’t proven to have any legitimate law-enforcement purpose. I think my customers would be resentful of having been the target of that kind of surveillance.”

Racial Bias
In 2012, the IEEE studied the detection accuracy of facial recognition algorithms performed on demographic groups divided by age, gender, and race/ethnicity. The study compared the results of commercial off the shelf, non-trainable, and trainable facial recognition algorithms. The detection accuracy of trainable algorithms can be improved for certain demographics by calibrating them with facial images of that demographic. Commercial off the shelf and non-trainable algorithms are unable to be customized.

Scans performed on female, black, and younger cohorts contained the most false-positives. Training face recognition systems with diverse data sets spanning all demographics reduces systematic bias for specific demographics. Training algorithms on specific demographics improves performance accuracy on that demographic. Multiple face recognition systems trained on a specific demographic can be used to reduce bias. The operator can select best algorithm based on given demographic information. This is known as dynamic face matcher selection.

ICE Policy
On December 22, 2017, Immigration and Customs Enforcement (ICE) was awarded "a contract to obtain query-based access to a commercially available License Plate Reader (LPR) database" with Vigilant Solutions. They are "neither seeking to build nor contribute to a national public or private LPR database," but will use it to aid in its investigations. The DHS Privacy Impact Assessment explains that "in some cases, when other leads have gone cold, the availability of commercial LPR data may be the only viable way to find a subject."

This contract has a lot of implications in enforcing immigration laws. In the first few weeks of Trump's administration, immigration arrests rose 32.6% and non-criminal arrests doubled. With access to LPR data, ICE investigators have the capacity to find and track any undocumented immigrants they have leads on, which could lead to arrest and possible deportation.

The database contains information on cars all over America that have driven in front of a camera linked to the network. In fear of this resource being abused, ICE has put in place regulations. These include mandatory training in non-discriminatory use of the system, permissible purpose before accessing the service, limited time frames for queries, regulated alert list that flags any of-interest license plates, and all queries are audit-able (Who? Why? When?).

DHS Privacy Impact Assessment
The department of homeland security did a privacy impact assessment prior to the contract between the ICE and Vigilant Solutions and came up with 6 concerns. The biggest concern is long-term aggregation that can lead to unwarranted surveillance and reveal details about people’s private lives. This can include places of worship, protests, and meetings which are all constitutionally protected freedoms. Additionally, there can be machine or user errors that lead to wrongful charges, there’s a limit to its tracking ability because it only tracks vehicles, information can be shared or held longer than necessary, and as technology advances, it can intrude on privacy even further.

Criminal Cases
Vigilant Solutions post criminal cases that were solved suing their LPR system to their website.

In one hit-and-run case, the victim caught only caught a partial of the plate. Using the LPR database, the officer was able to find and track the vehicle knowing that it was a gold minivan and the partial plate. The officer said, “The commercial license plate reader data in LEARN allowed me to solve this case, allowing the victim to receive insurance compensation and assisting the family of the elderly man to better understand the risks associated with him driving an automobile” .

In another hit-and-run case, the victim saw a gold Chevy Cavalier and caught a partial plate. In the same way, the officer found the vehicle using the LPR system and claimed that “Without the historical LPR data from Vigilant, this case would most likely have been closed as inactive and remain unsolved".

There was a kidnapping case where a woman was drugged during a blind date arranged through an online dating app. Her daughter noticed her missing and used the dating app to find the culprit. Using the license plate number the police had on file, they were able to locate the vehicle and rescue the victim. A representative of Vigilant Solutions said, “The license plate detection information allowed officers to locate the vehicle and the woman, who had been drugged and was not able to care for herself. The agency was able to quickly get her to safety”.

Conclusion
PLATESEARCH and FACESEARCH are both great technical solutions that help solve and reduce crimes within society. The technology has proven to solve, prevent, and deter crimes from occurring. Vigilant Solutions ensures their products follow all U.S. laws and regulations however, concerns over privacy are still present. As seen with the cases above, technology like facial recognition and license plate readers can be misused. Without the proper supervision and accountability of those who employ these technical solutions, Vigilant Solution's products lead to the unwarranted invasion of individual privacy. From racial targeting to invalid probable cause, we have cases where law enforcement agencies have time and time again abused their power by mishandling these products. For their technology to prove helpful to all of society, more training and proper check and balances have to be in place throughout those who implement Vigilant Solution's products.