Facial recognition technology is a great tool that’s highly useful. So why did San Francisco just ban the use of it by government agencies?
It’s Not Perfect
Facial recognition is hardly perfect and it’s not nearly as accurate as fingerprinting or iris scanning. This leads to the possibility these systems are more likely to implicate innocent people in crimes they didn’t commit.
It may be a matter of opinion, but if there’s one thing worse than not catching the bad guy, it’s convicting the wrong person for the crime someone else committed.
Equality & Fairness Concerns
In addition to the overall imperfectness of facial recognition, it’s also been proven that these technologies are worse at correctly identifying minorities. The most infamous example was in 2015 when Google’s algorithm was tagging African Americans in photos as gorillas.
Because of the higher error rate in identifying minorities, using facial recognition to identify criminals or generate leads on criminal activity is much more likely to result in a higher rate of innocent people of color being convicted for crimes they didn’t commit. Using facial recognition in its current state would skew the number of wrongful convictions of minority citizens.
While facial recognition has been used successfully to identify criminals, the potential for abuse is extreme. A while back I wrote about how the use of automated license plate scanners could be an invasion of privacy – and facial recognition technology is another prime candidate for such a risk.
Using facial recognition in public spaces would move us one-step closer to a surveillance state where data isn’t just kept on those suspected of wrongdoing, but on everyone just in case it’s needed in the future. If there’s one thing history has taught us, it’s that data like this will be misused when the opportunity comes.
San Francisco’s ban on the use of facial recognition technology doesn’t apply to private space or businesses, but it does attempt to put a halt on the use by government agencies. Until the technology is more accurate and less biased, it should probably remain that way.
If you found the information in this post to be useful, please consider subscribing up to our email list. There's no spam, we currently only send emails once a month or less.