A Policy Framework for Responsible Limits on Facial Recognition Use Case: Law Enforcement Investigations

The World Economic Forum, in collaboration with INTERPOL, UNICRI, and the Netherlands Police, has developed a governance framework to mitigate the human rights risks and technical errors associated with facial recognition technology (FRT) in policing. The framework emphasizes that FRT outputs must only be treated as investigative leads and must undergo human verification by trained experts. It provides nine core principles covering human rights, proportionality, accountability, system performance, and transparency, supported by a self-assessment tool for law enforcement agencies to evaluate their compliance and operational processes.

• Facial recognition technology (FRT) outputs in law enforcement must never be considered definitive identification but only as an "investigative lead," requiring manual verification by experts to prevent wrongful arrests.
• The framework establishes nine foundational principles for responsible FRT use, with "Respect for human and fundamental rights" serving as the overarching motivation for all other principles.
• To ensure accuracy and mitigate bias, the framework recommends that law enforcement agencies use standardized procurement processes and require vendors to submit algorithms to independent audits and large-scale tests, such as those conducted by NIST in the US.
• The framework prohibits the use of FRT for purposes other than biometric identification, recognition, or verification, specifically banning its use to infer ethnicity, gender, sex, age, emotion, opinion, health status, religion, sexual orientation, or for predictive analysis.
• Real-time FRT is identified as the most sensitive use case, requiring higher standards and additional safeguards, including authorization by an independent authority or the chain of command with a 24-hour notification requirement.
• A rigorous human-centric verification process is recommended, involving a manual analysis by a facial examiner followed by a blind peer review by two other experts to reach a consensus before a result is shared with investigators.
• Transparency is critical for public trust; agencies should publicly disclose their FRT definitions, the software version used, probe image procedures, and auditable records of search requests and results.
• A pilot exercise involving six law enforcement agencies from five countries revealed a significant lack of standardization in FRT procedures and highlighted critical gaps in officer training and public communication.

APA

World Economic Forum (2022). A Policy Framework for Responsible Limits on Facial Recognition Use Case: Law Enforcement Investigations. https://www3.weforum.org/docs/WEF_Facial_Recognition_for_Law_Enforcement_Investigations_2022.pdf

Chicago

World Economic Forum. A Policy Framework for Responsible Limits on Facial Recognition Use Case: Law Enforcement Investigations. 2022. https://www3.weforum.org/docs/WEF_Facial_Recognition_for_Law_Enforcement_Investigations_2022.pdf.

Wikipedia

{{cite report |author=World Economic Forum |title=A Policy Framework for Responsible Limits on Facial Recognition Use Case: Law Enforcement Investigations |date=November 2022 |url=https://www3.weforum.org/docs/WEF_Facial_Recognition_for_Law_Enforcement_Investigations_2022.pdf |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{worldeconomicforum2022policy, author = {{World Economic Forum}}, title = {{A Policy Framework for Responsible Limits on Facial Recognition Use Case: Law Enforcement Investigations}}, institution = {World Economic Forum}, year = {2022}, month = nov, url = {https://www3.weforum.org/docs/WEF_Facial_Recognition_for_Law_Enforcement_Investigations_2022.pdf}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }