Author(s): Raghu Gopal
Last week, San Francisco, one of the most tech-enthusiastic cities in the world, voted to ban the use of facial-recognition software by police and other municipal agencies. The move makes it the first US city to block a rapidly developing technology amid growing privacy concerns. Thanks to jurisdiction limits, the ban doesn't affect deployment by the federal government at San Francisco's international airport and seaport, nor does it limit personal or business use.
Facial recognition has improved dramatically in recent years, helped by the popularity of deep learning. In a typical system, facial features are analysed and compared with labelled faces in a database. Facial-recognition technology has helped law enforcement agencies around the globe to spot fraud and identify criminal suspects, but critics say that recent advances in artificial intelligence may have transformed the technology into a dangerous, Orwellian-like tool that enables real-time surveillance (see Smart City or Surveillance Hell?).
Studies by researchers at the Massachusetts Institute of Technology and Georgetown University have found that the technology is less accurate at identifying people of colour and women, and that it could automate biases already pervasive in law enforcement. For example, Amazon's facial-recognition tool incorrectly matched the faces of 28 lawmakers with people in mug shots, and disproportionately misidentified people of colour in a test by the American Civil Liberties Union.
The US Customs and Border Protection agency adopts facial recognition in many ports of entry. For example, at airports, international travellers stand before cameras and have their pictures matched against those provided in their passport applications. The agency says this complies with privacy laws, but the process has, nonetheless, been criticized by the Electronic Privacy Information Center, which argues that the government, despite promising travellers that they can opt out, has made it increasingly difficult for them to do so.
The ban on the technology, which is set to go into effect in a month, could become final after a second vote of the city's board of supervisors this week. Similar legislation is being proposed in the neighbouring city of Oakland. The state of Massachusetts is also weighing a bill that would postpone the deployment of facial-recognition software in the state until the technology improves.
Several high-profile stories about what could be judged as excessive use of tracking technologies in many countries, particularly China, could be leading to a biometrics backlash. But the genie is out of the bottle now, and this will be a difficult trend to reverse.