Is Facial Recognition Technology Capable of Identifying Suspects? Here’s What Tech Experts Say

San Francisco prohibited city police from using facial recognition technology to identify suspects, and now experts warn the fledgling big tech tool might not be ready for prime time.

Critics believe such technologies carry a huge risk as they are heavily guarded and the companies who deploy them tend to shy away from providing too much information about their use. One artificial intelligence researcher claims using facial recognition makes defending potential suspects nearly impossible.

“You can’t meaningfully build up a criminal defense, or change policies, if you don’t know how these tools are being used,” Alice Xiang, a researcher at the Partnership on AI, told an Axios reporter Thursday. There’s also significant discussion about whether facial recognition technology is capable of effectively identifying people once deployed.

Facial recognition technology is misidentifying people in London as potential criminals at a rapid rate, the Independent noted in a May 7 report. London carried out eight trials between 2016 and 2018 which resulted in a 96 percent false-positive rate, meaning the software wrongly alerted police that a person passing through the scan was a suspect, according to the report.

At least one American city is taking a proactive approach toward addressing concerns. San Francisco’s Board of Supervisors voted 8-1 Tuesday to block city police from relying on the tool to identify criminal suspects. Law authorities are not currently using the technology, media reports show.

Civil liberty groups worry the technology’s potential could push the U.S. in the direction of an overly oppressive surveillance state. They sometimes cite China’s massive data collections to justify their concerns.

The communist country, which is competing with the U.S. to deploy fifth generation mobile services, is deploying the technique to keep tabs on the country’s 11 million Muslims, many of whom are being held in internment camps, The New York Times reported Tuesday. China maintains a vast surveillance net in the western region of Xinjiang, yet the scope of the system extends beyond that small region.

Silicon Valley companies have also been under intense scrutiny for deploying technology some believe is disproportionately targeting conservatives. Tech experts have expressed concerns that Facebook and Google are not up to the task of distinguishing between legitimate content and hate speech. President Donald Trump and others have expressed concern that Twitter and Facebook are targeting them.

Emily Williams, a data scientist and founder of Whole Systems Enterprises, argues that Facebook’s lack of transparency about the frailties of their technologies feeds concerns. She told The Daily Caller News Foundation in April that Facebook’s algorithm has a 70 percent success rate, which means 30 percent of the time the tool is nixing conservatives who are sharing provocative but legitimate content.