Using technology to make us safer is smart. Pioneering the responsible use of that technology makes us a leader.
At the Sheriff’s Office we use technology to make our communities safer every day. We developed technology that provides timely arrest warrant information and other crime-related data to law enforcement officers at agencies across the county, helps detectives and Patrol deputies solve crime, and enables the display of important information in key areas in Jail and Records. It allows Jail deputies to log rounds in real time and in a manner which increases safety. Body camera technology improves evidence gathering, enhances safety, and increases transparency. We have been a leader in GPS-tracking technology to catch thieves who steal packages, cars, bicycles and more. We are also leaders in the responsible use of facial recognition technology.
It’s been stated that our Office was the first law enforcement agency to use facial recognition technology. This is not accurate. The truth is that the Sheriff’s Office did not use facial recognition until October 2017 when it was made available to investigators. A year earlier, in October 2016 the Georgetown Center on Privacy and Technology published an article citing 25 police agencies that used the technology, pre-dating our use. But while we were not one of the first agencies to use the technology, we do lead in developing model policy and responsible use of the technology to solve criminal cases.
So how did we become a leader in the responsible use of facial recognition technology?
As businesses and individuals increasingly used video technology and smart phones to capture suspects’ images, law enforcement had a corresponding opportunity and obligation to identify the unknown suspects caught on camera. Historically, this process was time-intensive and inefficient. An investigator would often create a bulletin and distribute it throughout their agency. Depending on the era, they would send it by fax or email to surrounding law enforcement agencies and jail staff as well, in hopes that someone might recognize the suspect.
In the mid-2000s, the Sheriff’s Office launched a “Can-You-ID Me?” page that allowed the public to view photos of suspects engaging in criminal activity and submit possible tips. The page proved to be a valuable tool to solve cases, and many law enforcement agencies followed suit. .
Fast forward to 2016, when personnel with the Sheriff’s Office Information Technology (IT) workgroup learned of a technology product announced by Amazon Web Services. The IT workgroup saw an opportunity to apply portions of this technology to assist in accomplishing the same mission as the “Can-You-ID Me?” page. Knowing the need to more quickly and efficiently identify criminal suspects, our IT developed a method to compare unidentified criminal suspects’ faces to Washington County booking photos. This quickly proved to be successful in assisting them in identifying suspects caught on camera.
We also understand that new technologies can raise new concerns for the public – which we share – including its appropriate and responsible use and potential for privacy violations.
We have worked hard to establish the policy (linked above) that guides our responsible use of facial recognition technology. Along the way we’ve consulted and collaborated with law-makers (Sen. Wyden’s staff) and local stakeholders for feedback that we’ve used to improve our policy and protocols.
Notably, our policy is clear that a facial recognition search alone is not probable cause to arrest or seize any person. Results of using the technology are only potential leads that require follow-up investigation, such as verification of the person’s true identity and evaluation of other information and facts. Sometimes such corroboration gets achieved through public social media sites (for example, the suspect’s clothing on social media matches what they wore during the crime).
Other important points:
- We do not use the technology for mass surveillance. Policy specifically prohibits such use.
- We only compare criminal suspect photos with Jail booking photos dating back to 2001, which are public record.
- Criminal suspect photos not resulting in a criminal case are purged after one year. This timeline is consistent with Oregon retention requirements and ensures all data is available for our annual audit of this technology’s use.
- All data collected is kept secure and only retained locally on Sheriff’s Office servers. No photos or videos are ever sent to the servers of private companies. Neither are they ever sent to law enforcement agencies based outside of Washington County.
In addition, we have been very transparent about this program. We’ve invited local television stations and a law enforcement technology publication to cover it. Between Facebook and Twitter accounts alone, stories on our use of facial recognition technology were seen 25,000 times. This does not include TV or web viewership of local TV stations’ broadcasts. These are some examples:
Using Facial Recognition to Fight Crime, KGW-TV Shared to Facebook (3.5K views) and Twitter (6.4K views)
Local deputies using facial recognition software to fight crime, KPTV-TV Shared to Facebook (4.2K views) and Twitter (4.2K views)
Washington County, Ore., Adds Facial Recognition to Suite of Investigative Tools, Government Technology News Shared to Facebook (3.9K views) and Twitter (3.3K views)
March 2018 Monthly Newsletter, WCSO Subscriber base of approximately 40,000
And while I believe we’ve developed a strong policy for the responsible use of facial recognition technology, I also strongly advocate our Oregon legislature establish a multi-disciplinary workgroup to determine limits on its use by law enforcement. I look forward to being part of that discussion, sharing our experience, and continuing to learn. In the meantime, we remain focused on being leaders in the responsible use of law enforcement technology as an important way to make our community safer.