Public Facial Recognition Moratorium vs Regulated Law Enforcement Use

A Public Facial Recognition Moratorium excels at preserving civil liberties and building public trust by establishing a clear, bright-line rule against government surveillance. This approach directly addresses documented concerns about the technology's higher error rates on women and people of color, as highlighted by the NIST Face Recognition Vendor Test (FRVT), which found false positive rates can be up to 100 times higher for certain demographics. Cities like San Francisco and Boston have adopted this model, effectively halting procurement of these systems and forcing a public conversation about acceptable use before deployment.

Regulated Law Enforcement Use takes a different approach by implementing strict guardrails—such as judicial oversight, mandatory algorithmic auditing, and human-in-the-loop requirements—to harness the technology's crime-solving potential while mitigating risks. For example, a 2019 study by the ACLU found that a single false match could implicate an innocent person, but proponents argue that regulated use with high confidence thresholds has successfully identified human trafficking victims and solved cold cases. This results in a trade-off where civil rights protections are procedural rather than absolute, relying on the strength of the auditing and oversight mechanisms.

The key trade-off: If your priority is eliminating the risk of biased misidentification and mass surveillance from the outset, choose a moratorium. If you prioritize providing law enforcement with a regulated tool for serious crime investigation, where efficacy is balanced by judicial checks and mandatory bias audits, choose a regulated use framework. The decision hinges on whether you trust preemptive prohibition or post-hoc oversight to protect fundamental rights.

A Public Moratorium excels at eliminating immediate civil liberty risks and building community trust because it provides a hard stop on a technology often criticized for racial bias and mass surveillance. For example, cities like San Francisco and Boston enacted bans after studies, such as the 2019 NIST report, found that many commercial algorithms were up to 100 times more likely to misidentify Black and Asian faces, making the moratorium a powerful tool for preventing discriminatory policing outcomes.

Regulated Law Enforcement Use takes a different approach by attempting to preserve the crime-solving efficacy of the technology while layering on judicial oversight and mandatory algorithmic auditing. This strategy results in a trade-off where agencies like the London Metropolitan Police have used live facial recognition to identify wanted individuals in crowds, but must continuously invest in auditor access, strict watchlist management, and transparency reporting to maintain operational legitimacy.

The key trade-off: If your priority is preventing any possibility of algorithmic discrimination and mass surveillance in public spaces, a moratorium is the definitive choice. If you prioritize a narrow, auditable tool for identifying known violent offenders under strict judicial warrants, a regulated framework with mandatory NIST-caliber bias testing and public transparency dashboards is the more surgical option. Consider a moratorium if your community trust is fractured; choose a regulated approach only when you can guarantee continuous, independent oversight and a zero-tolerance policy for general surveillance creep.