Why Big Tech's retreat from facial recognition isn't as altruistic as it looks

IBM and Amazon been applauded for moving out of a market many regard as immoral. But is this real change?

Sign Up

Get the New Statesman's Morning Call email.

As protests over police brutality and systemic racism spread around the world, business leaders have lined up to make corporate commitments to diversity and inclusion. But while many of these pledges have come from consumer brands, one of the most striking commitments has come from the US enterprise software giant IBM. 

Earlier this week, the company's chief executive Arvind Krishna wrote a public letter to the US Congress putting forward policy proposals to advance racial equality and announcing that his company would no longer sell general purpose facial recognition software, amid concerns that the fledgling technology could trigger a surge in discriminatory policing.

"IBM firmly opposes and will not condone uses of any technology, including facial recognition technology offered by other vendors, for mass surveillance, racial profiling, violations of basic human rights and freedoms, or any purpose which is not consistent with our values and Principles of Trust and Transparency," wrote Krishna. "We believe now is the time to begin a national dialogue on whether and how facial recognition technology should be employed by domestic law enforcement agencies."

The move has been welcomed by politicians and digital rights activists who have long warned that the technology, which regularly misidentifies women and people of colour, risks exacerbating discriminatory policy tactics and accelerating the end of privacy in public spaces.

However, analysts have cast doubt on just how big a sacrifice this is for IBM. 

"Computer vision as a service wasn't a particularly profitable business for IBM," says GlobalData analyst, Laura Petrone. In the facial recognition sector, which is just one subset of the wider computer vision market, IBM is reportedly trailing in third place.

"Big tech companies, especially those with access to large visual datasets, are the best positioned to benefit from selling the technology," explains Petrone. "Among big tech companies, Amazon and Google, which have heavily invested in ML [machine learning] and rely on advanced cloud operations, are the leaders." IBM, she says, "simply can't match those two companies' repositories of visual data."

Experts have also speculated that demand for the technology is likely to fall in the coming months if the technology becomes less reliable as the use of masks rises during the coronavirus crisis, making services less commercially viable. 

There is evidence, then, that this was at least as much a business decision as an ethical one. IBM's record does little to rebuke such an argument. The company has been criticised for reportedly developing video analytics that could search New Yorkers by skin colour; for bidding to work for the highly controversial US Immigration and Customs Enforcement agency; and for selling its Intelligent Operations Center technology, which offers "crime prevention and suppression" through video analytics, to Davao in the Philippines despite international condemnation of the city government's alleged complicity in huge numbers of targeted murders.

Given IBM's relatively small share of the facial recognition market, the impact of its decision on the adoption of the technology in law enforcement is expected to be limited. Microsoft had previously stated in 2018 that it wouldn't sell the technology to agencies using it for mass surveillance. However, the most significant consequence may be that it has now triggered Amazon, one of the market leaders, to issue a one-year moratorium on selling facial recognition technology to police forces.

In light of these decisions, the procurement of facial recognition technology by police forces and other law enforcement agencies will be even more controversial than it already was. But where there is a demand, it will ultimately be fulfilled unless there is regulation to prevent it. Politicians are the people responsible for controlling controversial technologies, not executives in Silicon Valley.

And while it is telling that San Francisco — the home of the US tech industry — has banned facial recognition, politicians in other states and countries remain conspicuously quiet on the matter. Until they act, decisions on how this technology should be adopted — if at all — will remain in the hands of local police forces and their technology providers, with or without IBM and Amazon.

Oscar Williams is the editor of the New Statesman's technology site, NS Tech.

Free trial CSS