San Francisco passed a bill to ban facial recognition yesterday – becoming the first major city to prohibit local govt agencies from using the technology.
“The most pervasive and risky surveillance technology of the 21st century.”
According to Dave Maass at Electronic Frontier Foundation, police depts in Vegas, NYC, Orlando, San Jose, San Diego, Boston, Detroit, and Durham are already using facial recognition technology. Overseas, the Chinese govt is known for using more than 200 million surveillance cameras to monitor the Uighur tribe.
Getting your every move tracked by cameras controlled by the police and govt agencies is in and of itself a major civil liberty concern. What makes it worse is that the facial recognition tech they use is deeply flawed. Alvaro Bedoya at Georgetown U’s Center on Privacy and Technology called facial recognition “the most pervasive and risky surveillance technology of the 21st century.”
Bias In / Bias Out
In computer science, there’s a phrase called “Garbage In / Garbage Out”, which describes the situation where flawed data input produces flawed outputs. The same problem is happening with facial recognition tech, in the form of racial and gender-bias.
Silicon Vally is dominated by white males who use datasets comprised of white males when developing machine learning algorithms. Naturally, the AI tech they come up with this fundamentally biased process end up being heavily biased.
Below are just a few examples of AI bias:
- In 2015, Google’s AI labeled black people as gorillas
- Last year, Amazon’s facial recognition showed 800% higher error rate for non-white members of Congress than their white counterparts
- Error rates for darker-skinned females in facial recognition testing conducted by MIT were 34.7%. For lighter-skinned males, the maximum error rate was less than 1%.
Basically, if you’re a non-white male, the police being equipped with facial recognition technology is pretty much a nightmare scenario.
What SF’s ban means
Despite all the glaring flaws of facial recognition tech, it’s still not regulated on the federal level – which prompted progressive cities like SF to take matters into their own hands.
The fact that San Francisco, a major tech hub, was the first city to pass the ban on facial recognition sends a strong message. Aaron Peskin, the city supervisor said that they “have an outsize responsibility to regulate the excesses of technology precisely because they are headquartered here.”
Other cities such as Oakland and Somerville, MA are set to pass similar bills and with SF setting the precedent yesterday, more cities are expected to follow their lead.