The Corner

National Security & Defense

Google, AI, and National Defense

Google CEO Larry Page speaks during a press announcement at Google’s headquarters in New York. (Eduardo Munoz/Reuters)

Last week, Dennis Prager castigated Google for its refusal to renew a contract with the Department of Defense that had arranged for the sharing of artificial intelligence for the analysis of drone footage. Prager’s point, that Google is abdicating a moral responsibility to do its part in the fight against evil, hits the mark.

Now that this initiative — “Project Maven” — has been abandoned, the big question is what Google will do next. Per Gizmodo, before caving to disgruntled Google employees and academics, the company’s executives had hoped to make the program palatable to the public by downplaying its military aspect. Google aimed to secure further defense contracts down the road, reportedly to the tune of a $10 billion cloud-computing agreement. Given recent events, one can see this distancing continuing. On issues ranging from gun rights to LGBTQ rights to health care, consumers today view political and social positioning as an inherent part of corporate branding. Google’s abandonment of Maven is the latest manifestation of this trend. Evidently, the company’s leaders decided that renewing the project would be less profitable than continuing it with vocal critics in hot pursuit.

Yet something tells me that Google can’t be totally counted out when it comes to cooperation with national-defense initiatives. While Google released a set of guidelines for its AI work Thursday afternoon that ruled out collaboration with governments to develop weapons, it said that it would continue working with the military in other domains. Silicon Valley and the Pentagon already both benefit from a substantial cross-pollination of investment and research and development. In the particular case of Maven, Google’s development of the geospatial technology critical to the project benefited significantly from special authorization it received from the government for access.

And this is pretty typical. The CIA’s venture capital fund In-Q-Tel is a key investor in several firms absorbed by America’s largest technology companies — think the tech that brought us Google Maps, and capabilities currently in development such as scanners to create 3-D printed objects. Eric Schmidt, the former Alphabet chairman, heads the Defense Innovation Board, perhaps the epitome of this cooperation.

There are structures and networks in place encouraging collaboration between government and industry. The next challenge is to harness them in a concerted way: With China, Russia, and France, among others, developing national strategies for AI, the United States lags behind. And as Vladimir Putin has put it, whoever masters artificial intelligence will “become the ruler of the world.”

That sounds dramatic, but it’s not wrong. In the latest issue of The Atlantic, Henry Kissinger describes the stakes in no uncertain terms in an essay titled “How the Enlightenment Ends.” He calls for a presidential commission on artificial intelligence to investigate its implications, else technology will progress without a corresponding and necessary advancement in our institutions. Those tasked with keeping America safe — with help from partners in Silicon Valley — will need to embrace their key role in this developing debate. While Maven’s expiration and Google’s new AI guidelines are a setback for national security, neither side should let these incidents, nor the trend that they represent, obstruct future cooperation.


The Latest