Google will not be renewing its drone contract in 2019 – called Project Maven – with the Department of Defense after pressure from its workforce, making it an unlikely move in a longstanding relationship between big tech and the military.
You can edit or expand this storyEdit
Project Maven, known technically as Algorithmic Warfare Cross-Functional Team, was an effort to improve the Pentagon’s artificial intelligence capabilities, created in collaboration with the U.S. military’s Defense Advanced Research Agency (DARPA).
In the wake of the internal revolt at Google, the company has released its latest guidelines regarding AI, pledging not to use its technology for weapons or illegal surveillances. However, the company said it will keep working with the military in other areas.
“How AI is developed and used will have a significant impact on society for many years to come,” wrote Google CEO Sundar Pichai in an accompanying blog post. “As a leader in AI, we feel a special responsibility to get this right.”
A growing concern over AI
Project Maven was controversial in large part because of the AI aspect. When Google employees wrote their open letter to Pichai, they noted the “growing fears of biased and weaponized AI” as a strong reason for their revolt.
Concerns over AI have been growing with outspoken critics such as Elon Musk voicing its potential threat, though much of the concern is partially exacerbated by sensational media headlines about the impending doom of AI.
Anxiety over autonomous weapons and killer robots is also growing, with many all over the world becoming more vocal about their existence. Project Maven aside, elsewhere this year big tech has been criticized for its relationship with the defense industry.
In April, AI experts boycotted a South Korean university for its partnership with a leading defense company. In May, civil rights groups criticized Amazon for working with the police to provide facial recognition technology.
In the face of this, a growing number of research organizations, such as OpenAI and the Future of Humanity Institute, are looking into how to make artificial intelligence more ethical to counteract the existential risk AI could bring if not developed safely.
Silicon military history
There is a longstanding history between Silicon Valley companies and the defense sector. Journalist Yasha Levine, in his book Surveillance Valley, traces this relationship by looking at, in particular, Google’s role as a defense contractor.
It is hard to tell where [these companies] end and the U.S. government begins. – Yasha Levine
In a section for left-leaning publication The Baffler, Levine wrote of Google:
“Over the years, it had supplied mapping technology used by the U.S. Army in Iraq, hosted data for the Central Intelligence Agency, indexed the National Security Agency’s vast intelligence databases, built military robots, colaunched a spy satellite with the Pentagon, and leased its cloud computing platform to help police departments predict crime.”
Google, of course, isn’t the only big tech company involved in the business of war.
“From Amazon to eBay to Facebook — most of the internet companies we use every day have also grown into powerful corporations that track and profile their users while pursuing partnerships and business relationships with major U.S. military and intelligence agencies,” wrote Levine. “Some parts of these companies are so thoroughly intertwined with America’s security services that it is hard to tell where they end and the U.S. government begins.”
Through this lens, it’s easy to assume that big tech permeates the defense industry, especially when some of humanity’s greatest achievements, notably the internet, come from the marriage between technology and the military.
When looking at the “total defense industry package,” said Galbreath, Google’s involvement is “more of a flirtation.”
So why is there such a misconception that this relationship is actually stronger than it is?
“I think it has something to do with the wow factor,” said Galbreath.
“Everybody knows what Google is, and the second thing [is] everybody’s interested in robotics and machine learning, etc. Thirdly, I suspect because it was ironic that you had this company that started out with this notion of “do no harm,” only then to turn around and be seen as trying to push a military revolution in terms of robotics.”
“Those are the three things that really make this notion of big tech and the military look like it is, in a sense, kind of cutting-edge and something the military is wholeheartedly investing in, and why big tech would be the perfect partner for that.”
Most tech in the military is still hardware
The U.S. military’s Defense Advanced Research Agency’s (DARPA) work with AI and machine learning is only one portion of what it does. “A lot of militaries are particularly interested in material sciences …[such as] armoury. It’s the very traditional companies that are making armour, not the tech companies such as Google,” Galbreath told WikiTribune.
Traditional companies include BAE Systems, Boeing and McDonnell Douglas. Most makers of drones, for example, are technically tech companies, but they’re not tech companies in the sense of the civilian world: they don’t produce products that permeate everyday life.
The nature of Google’s business is perhaps why the Maven workers succeeded in their protest. The business model of big tech in the civilian world is different from the business model in companies such as BAE, where the Google achievement probably couldn’t be replicated.
“The defence contractors are pretty impervious to critique, as we can see going back many, many years. And also, there’s a very close relationship between the Pentagon or those contractors and the UK defence establishment. They’re more impervious to those things,” said Galbreath.
Google will quietly, cautiously continue military work
Levine, the author of Surveillance Valley, said that Google pulling out Project Maven was essentially a PR move.
I love how everyone is falling for Google’s PR move. Sure, it might not review this specific AI drone contract. But what about the rest of the company’s military contracting work? What about it’s work with predictive policing outfits? https://t.co/8mgAeotfEO
Despite dropping out, Google is still competing with other tech giants, including Microsoft Corp., IBM and Oracle, to pursue the Joint Enterprise Defense Infrastructure (JEDI) cloud contract from the Defense Department, which is worth more than $10 billion (€8.5bn) over ten years. This is echoed in a piece by Motherboard, which argued that quitting its Maven contract doesn’t mean Google is leaving the business of war.
Google’s goals, fundamentally, are commercial, not military.
“I’m very sceptical about Google. Because I think it makes a lot of sense for Google to be shown to having a lot of employee power. I’m not sure that I really buy the Google line that we have these hacktivist employees who in some way after campaigning for years and years have finally overturned some kind of policy,” said Galbreath.
“I would even go so far to say one of the reasons it was easy for Google to say that the employees were powerful in overturning a … company policy is because increasingly Google was becoming irrelevant to what the military was doing.”
That’s because Google’s goals, fundamentally, are commercial, not military. Google is a tech company that works primarily in the civilian world: its products sit in people’s homes, on their phones, and are used on a regular basis. This means that as it becomes more infused into everyday life, the more it wants to make sure it appears responsible and ethical.
Google won’t be ending its military contract, but because of its image, it has to be more careful in how it handles future involvements. As its employees wrote in their open letter to Google’s CEO, “its direct reach into the lives of billions of users set it apart.”
- See other WikiTribune coverage on AI and ethical use