Detroit’s new facial recognition policies: a potential model for the nation?
This year, the Detroit Police Department (DPD) has undertaken significant changes to its facial recognition technology (FRT) policies, in the wake of an ACLU-supported lawsuit following wrongful arrests based on FRT errors.
The use of FRT in law enforcement has been a contentious issue due to its potential for misuse and the risk of wrongful arrests.
In Detroit, this concern became a reality with several high-profile cases.
In 2020, Robert Williams was arrested after being misidentified by FRT. He spent 30 hours in jail before being released and subsequently sued the DPD. This case highlighted the serious implications of relying solely on FRT for arrests, particularly for people of color, who are disproportionately affected by misidentifications.
Similarly, in 2023, Porcha Woodruff was wrongfully arrested while eight months pregnant, due to an FRT misidentification. Her case further underscored the technology's limitations and the urgent need for reform.
Williams' lawsuit, supported by the ACLU, resulted in a groundbreaking settlement. The settlement includes new policies aimed at preventing future wrongful arrests and a significant step towards more responsible use of FRT.
The new policies emphasize accountability, transparency, and safeguards, potentially transforming Detroit into a national model for the ethical use of FRT in law enforcement.
Facial recognition technology in law enforcement: technology’s evolving role in policing
FRT can enhance public safety and facilitate suspect identification. It is also useful in locating missing persons and identifying witnesses. Many US police departments utilize FRT for these purposes, benefiting from its ability to quickly process large volumes of data.
However, FRT in law enforcement is not without its challenges.
Biases related to race, age, and gender can result in higher rates of misidentification among certain groups. Issues with low-quality surveillance footage further exacerbate these issues, leading to false positives.
There has been considerable pushback against FRT in various states. In 2020, several states imposed moratoriums on its use by law enforcement. San Francisco banned it entirely in 2019, and companies like Microsoft have restricted police departments from using their AI technology for facial recognition.
Detroit’s new policies stand out as a proactive response to these concerns, setting the stage for a more balanced and ethical approach to FRT.
Key changes in Detroit's new FRT policies and their potential impact
The new policies in Detroit include several critical measures designed to address these challenges:
Prohibition of arrests based solely on FRT results: this ensures that FRT is used as a tool to support investigations rather than the sole basis for arrests.
Requirement of independent, reliable evidence: any photo lineup based on FRT must be supported by independent and reliable evidence, adding a layer of verification to prevent wrongful arrests.
Mandatory audits: all FRT-related cases since 2017 will be audited. This is a positive step towards accountability, ensuring past mistakes are identified and corrected.
Training for officers: officers will receive training on the risks and limitations of FRT, which is crucial for minimizing errors and understanding the technology's impact on different demographic groups.
These policy changes address many safeguards associated with responsible FRT use and emphasize transparency, accountability, and rigorous oversight. The policies also advocate for responsible use of FRT with clear guidelines and robust training, alongside continuous monitoring and evaluation of the technology's effectiveness and fairness. They also integrate many of the data privacy principles in policies like the GDPR and the upcoming EU AI Act.
How will this policy change impact the wider US law enforcement landscape?
Detroit’s policies stand out not just because they address the challenges associated with FRT but because they establish a comprehensive system of checks and balances. By introducing stringent controls, Detroit emphasizes that FRT should not be a standalone solution, but part of a broader investigative toolkit that requires independent verification and careful oversight.
The city’s policies are rooted in ongoing audits, officer training, and strict evidence requirements, which can serve as a blueprint for departments across the country that want to integrate FRT responsibly.
Moreover, Detroit is actively fostering public trust in its use of technology by aligning its policies with national and global standards for privacy and transparency. These reforms suggest that when used responsibly, FRT can significantly enhance public safety without compromising ethical standards—a lesson that could reshape how law enforcement agencies across the US implement FRT in the future.
While these policies represent significant progress, their long-term effectiveness will depend on consistent application and periodic review. The ongoing dialogue and collaboration among stakeholders will be vital to ensure that the use of FRT remains ethical, fair, and proportionate.
While FRT can be a powerful tool for enhancing public safety, it must be used responsibly and with robust safeguards to protect individual rights. Detroit's experience can offer valuable lessons for other jurisdictions aiming to implement FRT in a fair and accountable manner.