During the second world war, IBM supplied the Nazis with technology used to help transport millions of people to their deaths in the concentration camps.
The American technology company leased punch-card machines through a German subsidiary for the purpose of tabulating a population census, which allowed the Nazis to identify and track the movements of Jews all the way to the gas chambers. So helpful were IBM’s machines that Hitler awarded a special medal to the company’s CEO, Thomas Watson, in recognition of his services to the Third Reich.
This shameful episode in IBM’s history highlights an uncomfortable predicament for companies pursuing lucrative government contracts, and one that’s recently reared its head for Google, Microsoft and Amazon: should they be held accountable for how their customers deploy their technologies?
The question has come to the fore for Microsoft, after it was revealed that the company’s Azure cloud computing arm was working with US Immigration and Customs Enforcement (Ice).
The relationship was announced in January, when Microsoft said it was “proud to support” Ice’s work with secure cloud services and facial recognition and identification. However, it came under scrutiny this month as the ugly results of Trump’s zero tolerance border policy, which saw thousands of children separated from their parents and detained in cages, unfolded.
For Microsoft’s executives, the deal marks an important expansion into a lucrative government market. Employees saw it differently, and hundreds signed an open letter to Microsoft CEO’s Satya Nadella demanding the company end its $19.4m contract with Ice and create a policy not to work with clients who violate international human rights law.
“We believe that Microsoft must take an ethical stand, and put children and families above profits,” they wrote. “As the people who build the technologies that Microsoft profits from, we refuse to be complicit.”
It’s easy to think of “the cloud” as a neutral bank of computers to which Ice has bought off-the-shelf access in the same way any business can buy cloud storage, laptops, smartphones or accounting software. We don’t blame Apple if an Ice agent is tipped off about an undocumented worker via a call to his iPhone, so what’s the big deal with Microsoft selling cloud services to the agency?
“The Microsoft cloud that the government uses looks really different from the cloud from [consumer service] BlueHost,” said Lilly Irani, an assistant professor of communication and science studies at UC San Diego, told the Guardian.
Serving the specific needs of a client like Ice, including supporting facial recognition and real-time translation of voice audio, would require a highly customised infrastructure with state-of-the-art data processing abilities, she explained. This means Microsoft engineers dedicating their time to build and maintain tools that help catch and deport immigrants.
Even if there were no customisation, Microsoft is complicit, argues Nicholas Evans, an assistant philosophy professor at UMass Lowell, who has studied the complicity of psychologists in torture programmes.
“By choosing to take Ice’s business they are saying ‘we think Ice isn’t objectionable enough that we won’t do business with them’,” he said. “This tacit endorsement is problematic for Microsoft.”
This is the latest in a series of employee revolts within Silicon Valley.
In April, Gizmodo revealed that Google had partnered with the US Department of Defense (DoD) to help the agency develop artificial intelligence for military drone surveillance through a pilot project for the Pentagon’s Project Maven, which seeks to identify objects in drone footage.
Staff who had joined a company famed for its search engine and advertising business were horrified to find themselves providing services to the DoD. It prompted a huge internal backlash and dozens of employees resigned.
The worker action was effective: Google announced it would not renew its Project Maven contract.
Amazon also came under fire for selling a facial surveillance tool “Rekognition”, which can identify, track and analyze people in real time, to various law enforcement agencies. According to Amazon’s marketing material the software can recognise “persons of interest” from a police body-cam video stream.
“Technology companies should not knowingly provide dangerous surveillance companies to government,” said Matt Cagle from the ACLU, which first highlighted the problems with Amazon’s facial recognition technology. “The marketing materials for [Rekognition] would make George Orwell blush.”
“It’s disingenuous for these companies to abdicate the responsibility for harmful use of their products,” he added.
As with Google and Microsoft, the backlash triggered workers to write a letter to Amazon’s CEO Jeff Bezos, obtained by Gizmodo, stating: “Our company should not be in the surveillance business; we should not be in the policing business; we should not be in the business of supporting those who monitor and oppress marginalized populations”.
The backlash marks something of a turning point for Silicon Valley’s tech workers, who have long been lambasted for their lack of empathy for those living outside of their gilded bubble.
“It’s taken a bit of effort to get them thinking about the impact of their work and be more aligned with the working class and immigrant populations as opposed to their executives and bosses,” said Ares Geovanos from the Tech Workers Coalition, which advocates for worker rights.
As we saw with Google, collective worker action can be particularly effective in companies whose bottom lines rely on their ability to retain highly-skilled computer scientists.
“Those engineers collectively have more power than individual consumers to push their companies to be responsible corporate citizens,” said Evans.
Contact the author: firstname.lastname@example.org