Last week, three tech giants announced their decisions to back away from the facial recognition business. On Monday IBM announced that the company will stop developing or selling facial recognition software, calling for a “national dialogue” on whether it should be used at all. On Wednesday Amazon paused police use of its facial recognition for a year, hoping the moratorium “might give Congress enough time to put in place appropriate rules” for the technology. On Thursday Microsoft announced that the company did not sell facial recognition to the police, and would not until the government passes federal legislation regulating the technology.

How big is the facial recognition global market? According to Market Research Future's report published in 2018, the global market for facial recognition is set to touch $8 billion in 2022 up from $3.04 billion in 2016, expanding at a compound annual growth rate of 19.68%. IBM critics have pointed out it was hardly a great sacrifice for IBM to quit a market that it barely had a foothold in to begin with. Compared with IBM and Microsoft, Amazon is a prominent supplier of facial recognition software to law enforcement. It was a striking change for Amazon. Amazon Rekognition (facial recognition with AI, created by AWS) is used by police departments all over the United States. In an interview with Andy Jassy, CEO of AWS, when asked how many police departments across the U.S. are using Amazon's facial recognition tech, Jassy said, "I don't think we know the total number of police departments that are using [Amazon's] facial recognition technology. We have 165 services in our technology infrastructure platform, and you can use them in any combination you want." 

Back Away From Facial Recognition

Close

Back Away From Facial Recognition

Face recognition technology is increasingly used by law enforcement and at airports. As shown in the Perpetual Line-up research conducted by Georgetown University in 2016, law enforcement face recognition networks include over 117 million American adults. That is to say, about one in two American adults is in a law enforcement face recognition network. Major police departments are exploring face recognition on live surveillance video, but there is no independent testing regime for racially biased error rates. The announcements from Amazon and Microsoft apply only to police use of the facial recognition product. Neither Amazon nor Microsoft mentioned whether it has sold or will sell product to other federal agencies, such as FBI, ICE (U.S. Immigration and Customs Enforcement), or the Department of Defense. 

What are the consequences from the tech giants' decisions to pause or exit the facial recognition field? It remains unclear whether it will be a one-off change or a move that helps to nudge more tech companies into action. The main remaining companies focusing of facial recognition in the US include Google. But racial bias is not a new concern for Google. In April 2020, Google apologized after its Vision AI produced racial biased results, as the image of a dark-skinned hand holding a thermometer was labelled “gun” and the same image with a salmon-colored overlay on the hand was enough for the computer to label it “monocular”. Google since updated its algorithm and it does not return a “gun” label anymore as of April 6th. Back in 2015, Google declared itself "appalled and genuinely sorry" after a black software engineer tweeted that the Google Photos image recognition algorithms labeled photos of him with a black friend as “gorillas”. “Gorilla,” “chimp,” “chimpanzee,” and “monkey” were censored from searches and image tags in Google Photos after the 2015 incident.

Meanwhile facial recognition has much wider adoption in China and shows no sign of slowing down. Facial recognition is widely used in security surveillance, traffic control, online banking, and access to office buildings, as the racial bias and privacy issue were less a concern in China. The competitive landscape in facial recognition include new computer vision companies (SenseTime, Megvii, Yitu, and CloudWalk) and traditional surveillance providers (HikVision, Dahua, and UNView). Tech Giant Tencent began testing facial recognition in games as part of anti-addiction measures for minors, and a total of 80 games under Tencent will implement new anti-addiction regulations by the end of May 2020. Among the main competitors in the facial recognition field in China, SenseTime, Megvii, Yiyu, CloudWalk, HikeVition, and Dahua are all on the Entity List by the U.S. Department of Commerce. It means the companies are subject to specific license requirements for the export of certain items, and any transaction with the company listed on the Entity List is seen as a “red flag” by the U.S. government. The facial recognition products from China will not be able to enter the U.S. market, while the U.S. tech giants are backing away from facial recognition. It is uncertain if it will create a gap for smaller facial recognition companies in the U.S. to take over the market share given the temporary lack of tech giants and regulations in the facial recognition field.