Over the past decade, facial recognition software has emerged as one of the most powerful biometric technologies capable of identifying and verifying individuals based on a digital image or a video frame. Improvements in facial recognition systems – due mainly to progress in machine learning – are expected to boost the market for this technology to $15 billion in 2025, compared with $4 billion in 2017.
Facial recognition technology has many applications, from identifying rare genetic disorders and improving consumer experiences in the banking and retail sectors to speeding up security checks at airports. While the development of this technology creates unprecedented opportunities for socially beneficial uses, it also poses a serious threat to human rights and civil liberties – most notably the freedom of expression, the freedom of assembly and association, the right to privacy and the right to a fair trial. Furthermore, despite significant progress in recent years, studies have shown facial recognition to be biased, performing more poorly on people with darker skin tones and on women. In addition, some applications are particularly worrying, such as determining sexual orientation based on facial scans.
Governments must act to ensure the fair and transparent use of facial recognition systems by public and private institutions alike. Actionable policies can ensure the safeguarding of our fundamental rights, guide the socially beneficial development of facial recognition technology and foster public trust.
This project convenes stakeholders from all sectors of society to co-design guidelines for:
– Ensuring that government use of facial recognition technology does not infringe on individual rights and fundamental freedoms
– Preventing and mitigating uses that are likely to lead to biased outcomes
– Protecting consumers against invasive surveillance by private companies