Facial recognition technology (FRT) is a type of biometric recognition technology that uses artificial intelligence (AI) to identify individuals through their facial features. FRT functions by extracting biometric information based on key facial features and makes comparisons between live and stored biometric templates. Facial recognition software produces recommended matches when it identifies a level of similarity between images that exceeds a set confidence threshold.
Defining Facial Recognition Technology
In Canada, facial images are considered to be “highly sensitive” biometric information because they are “a unique and permanent characteristic of our body, largely stable over time, and a key to our identity.” As such, they require a higher standard of data protection.
Facial recognition technologies often operate with deep biases as their algorithms’ choice architecture reflects the biases and values of their creators, and of the datasets on which they are trained. There is also evidence to suggest that the way cameras capture images of darker-skinned faces further undermines FRT’s efficacy.
As facial recognition software is better trained to identify individuals with White and/or White passing features, its use creates increased opportunities for misidentification for members of equity-deserving communities including Indigenous, Black, and racialized individuals, as well as women. Given the absence of regulation and the low barriers to entry into the FRT market, there are a growing number of facial recognition algorithms of varying quality operating in Canada.
Currently there are no industry wide standards for these algorithms’ confidence thresholds, further increasing the risk of false identifications. Worryingly, even if FRT systems rendered perfectly accurate analyses, they would still contribute to mass surveillance as they can be deployed to contribute to the over-surveillance and over-policing of equity deserving populations.
Absent explicit legislative direction, public and private entities in Canada, including public safety authorities and retail corporations, have regularly violated individuals’ privacy rights through their use of FRT. Its widespread accuracy issues pose heightened risks to equity-deserving communities. Its use in public spaces online or off undermines people’s ability to gather anonymously, right to freedom of expression, and contributes to mass surveillance. Regulation is needed to clearly define the context in which FRT could potentially be used, and to ensure the laws that govern its use safeguard individuals’ rights, given the sensitivity of the information and the risks of harmful impacts.
Legal Status of FRT in Canada
Today, FRT is governed by a patchwork of dated legislation, including federal and provincial public and private sector privacy laws, and the Charter of Rights and Freedoms. There is a lack of specific and comprehensive national legislation to direct its use in the private and public sectors.
While the law has remained stagnant, FRT is increasingly being adopted across many sectors. Absent legislative direction, courts have reached varying conclusions on issues surrounding its deployment and use by public safety authorities. Governments in Canada, at both the federal and provincial levels, must take action to ensure that the rights of Canadians are safeguarded vis-à-vis this emerging technology.