Even accurate and unbiased facial recognition brings privacy, cost harms, groups say

Demands to halt facial recognition use in Canada, at least pending further study and legislation, are not being met, or at least not quickly enough for some. A letter of concern has been jointly signed by The Canadian Civil Liberties Association and a list of signatories including representatives of civil society groups and academia.

A subtle change in the way the risks of facial recognition are described may be the most noteworthy element of the letter to biometrics developers, particularly outside of Canada. Part of what makes it noteworthy, is the similarity of the shift to that of other advocacy groups fighting against the technology’s use.

A boycott of venues using facial recognition by several popular music stars and reported by Rolling Stone has been organized by Fight for the Future.

Many of the musicians were already signatories to a campaign launched by Fight for the Future against the use of facial recognition at concert venues in 2019.

Facial recognition was used by South Wales Police to look for wanted individuals amongst crowds of Harry Styles fans ahead of a concert in Cardiff last week, which drew the ire of Big Brother Watch.

Familiar campaigns against such deployments invoke the same reasoning as in previous years, but as indications of higher accuracy and less demographic bias than some opponents suggest arise, the arguments made to the public and public representatives are evolving.

A response by the Canadian government to a set of recommendations on regulating facial recognition by a parliamentary committee are inadequate, the letter says. The government said in a February 2023 letter that it is reviewing the country’s privacy legislation, and is open to a new framework for facial recognition.

The signatories of the CCLA letter want a recommended moratorium to be enacted without delay, and a study into the harms of biased or inaccurate algorithms carried out.

“However, even if increased accuracy is achieved within FRT applications, and technical bias is resolved, we note that the technology still raises serious concerns,” the CCLA writes. “If left unregulated, the use of more accurate facial recognition technologies will become even more detrimental to groups that already experience systemic discrimination.”

Similarly, Fight for the Future argues that more effective algorithms will actually do more harm.

“Even scarier, though, is a world in which all facial recognition technology works 100 percent perfectly – in other words, a world in which privacy is nonexistent, where we’re identified, watched, and surveilled everywhere we go.”

The collection of face biometrics by police comes with cost to the force, and to the privacy rights of members of the public, argues Big Brother Watch.

“At South Wales Police’s last deployment, over 86,000 innocent concertgoers were scanned with zero matches,” the group notes in comments to the BBC.

If the groups are able to establish, either legally or in popular opinion, that those whose photographs are analyzed with biometric algorithms are being harmed, it would likely be the strongest argument yet in favor of restrictions of facial recognition use.

Article Topics

accuracy  |  biometric bias  |  biometrics  |  data privacy  |  facial recognition  |  regulation