Edmontonians Face AI Surveillance as Police Trial Bodycams

The Edmonton Police Service (EPS) has announced a controversial plan to implement artificial intelligence (AI) facial recognition technology on officers’ body-worn cameras, raising significant concerns among residents. This initiative, touted as a world-first pilot in partnership with Axon, an American body-camera manufacturer, is set to redefine the role of these cameras from instruments of accountability to tools of mass surveillance.

Originally designed to enhance police transparency, body-worn cameras were meant to document interactions between law enforcement and the public. They aimed to create an objective record, deter misconduct, and foster trust within the community. However, the EPS appears to be shifting this foundational purpose. Instead of primarily monitoring police conduct, the cameras will now enable officers to surveil the public, effectively treating community members as potential suspects.

This “proof of concept” is scheduled to run through December 2023, during which the system will assess its ability to identify individuals with outstanding warrants or other safety flags. Operating in “silent mode,” the technology will not provide live alerts to officers. Instead, footage will be reviewed later for algorithmic matches against police databases, raising ethical concerns about privacy and consent.

Edmontonians may unknowingly be subjected to scans while going about their daily lives, their faces compared to criminal databases without their knowledge or approval. This raises alarm for many, as the public is being transformed into a pool of potential suspects, undermining the very trust that body-worn cameras were intended to build.

Global Testing Ground for Controversial Technology

The decision to utilize Edmonton as a testing ground for Axon’s facial recognition technology is unprecedented. No other police service globally has deployed such technology on body-worn cameras. This move places Edmontonians in a position where they may be unwitting participants in a trial for a product that Axon could market to other jurisdictions in the future.

This initiative contradicts Axon’s own prior commitments. In 2019, the company’s independent AI Ethics Board deemed facial recognition technology too flawed and biased for ethical deployment on body-worn cameras, recommending against its use. Axon had publicly promised to refrain from integrating facial recognition into its cameras, making the current trial perplexing and fraught with potential risks.

The implications of real-time identification and tracking by police are profound. If law enforcement can monitor individuals continuously, it could create a chilling effect on fundamental rights. Citizens may hesitate to participate in lawful protests or frequent certain public spaces, knowing their identities could be captured and stored in a police database.

Impact on Civil Liberties and Communities

The situation raises critical questions about the protection of rights under the Canadian Charter of Rights and Freedoms. Section 8 guarantees individuals protection against unreasonable search and seizure. The practice of running facial scans on everyone without suspicion, warrants, or consent appears to conflict with this protection. Furthermore, Section 7 ensures life, liberty, and security of the person, while Section 15 promotes equality. The potential for constant biometric monitoring disproportionately affects racialized and marginalized communities, exacerbating existing inequalities.

Surveillance technologies are not impartial. Studies have shown that facial-recognition systems exhibit higher error rates for women and individuals of color. The deployment of such a system in Edmonton could lead to unjust outcomes, including more false accusations or unwarranted stops for racialized residents.

The legal and social ramifications of mass biometric surveillance stretch the boundaries of constitutional rights and privacy expectations. The Canadian courts have yet to provide definitive rulings on these issues. Given Axon’s previous report criticizing the deployment of facial recognition technology, it raises concerns whether these systems have become sufficiently accurate and unbiased to justify their current rollout.

Despite advancements in technology, facial recognition remains deeply flawed, performing well only in controlled settings with clear images of well-represented groups. In real-world scenarios, such as those involving moving officers, crowded streets, and poor lighting, the reliability of these systems diminishes significantly.

As Edmonton embarks on this experimental journey, the community deserves a choice regarding their participation. An open debate on the implications of integrating AI surveillance into policing is essential before the EPS transforms accountability cameras into pervasive identification tools. The residents of Edmonton should not be the first to bear the legal and social costs of this untested technology.