Responsible Limits on Facial Recognition Use Case: Flow Management Part II

The World Economic Forum presents a multi-step methodology to operationalize the responsible use of facial recognition technology (FRT) in 'flow management' scenarios, such as airport boarding. The document introduces a quality management system audit framework co-designed with AFNOR Certification and a corresponding certification scheme. It features a case study of Narita International Airport's 'One ID' programme, which used the framework for self-assessment to demonstrate transparency and risk mitigation regarding bias and data protection.

• The proposed governance framework for FRT follows a four-step process: defining principles for action, designing best practices, assessing compliance via a questionnaire, and validating compliance through an audit framework and certification scheme.
• The audit framework focuses on a quality management system audit rather than auditing the FRT algorithms themselves, as the goal is to ensure the quality of service and the rigour of human organizational procedures.
• The current version of the principles for action consists of 10 core principles: proportional use, risk assessment, bias and discrimination, privacy by design, performance, right to information, consent, information display, right of access to vulnerable groups, and alternative option and human presence.
• Narita International Airport implemented the 'One ID' programme to streamline the passenger journey from check-in to boarding, utilizing FRT to allow passengers to move at a 'walking pace'.
• To mitigate bias in its FRT system, Narita International Airport selected NEC Corporation after NIST studies indicated its solution was among the least biased across age, gender, and ethnicity.
• Narita International Airport manages privacy risks by following a multistakeholder guidebook and deleting biometric data collected at check-ins within 24 hours.
• The certification scheme is voluntary and intended for public or private organizations using FRT for flow management; it is delivered by independent third parties operating to ISO/IEC 17021-1:2015 standards.
• The certification process involves a documentary audit of design and implementation, followed by a field audit of the system in operation at a date unknown to the applicant organization.
• Major non-compliance during the certification process results in the withdrawal of the certificate and a requirement to stop communicating the certification to the public.

APA

World Economic Forum (2020). Responsible Limits on Facial Recognition Use Case: Flow Management Part II. http://www3.weforum.org/docs/WEF_Responsible_Limits_on_Facial_Recognition_2020.pdf

Chicago

World Economic Forum. Responsible Limits on Facial Recognition Use Case: Flow Management Part II. 2020. http://www3.weforum.org/docs/WEF_Responsible_Limits_on_Facial_Recognition_2020.pdf.

Wikipedia

{{cite report |author=World Economic Forum |title=Responsible Limits on Facial Recognition Use Case: Flow Management Part II |date=December 2020 |url=http://www3.weforum.org/docs/WEF_Responsible_Limits_on_Facial_Recognition_2020.pdf |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{worldeconomicforum2020responsible, author = {{World Economic Forum}}, title = {{Responsible Limits on Facial Recognition Use Case: Flow Management Part II}}, institution = {World Economic Forum}, year = {2020}, month = dec, url = {http://www3.weforum.org/docs/WEF_Responsible_Limits_on_Facial_Recognition_2020.pdf}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }