This Group Pushed More AI in US Security—and Boosted Big Tech

In the highly competitive cloud computing market, Amazon, Google, Microsoft, and Microsoft are archenemies. In late 2018, the top executives of the four companies, including the future CEO of Amazon Andy Jassy joined forces to do an unpaid side job: advise the president and the US Congress about how artificial intelligence could boost national security.
The National Security Commission on AI was created by Congress and named the executives. Eric Schmidt, the former CEO of Google, was its chair. He later stated that it would allow the US to harness this transformative technology in order to benefit our national and economic security.

Schmidt, Jassy and other members of the Big Tech commission had an economic interest in this topic. Their companies compete for Pentagon contracts such as the $10 billion JEDI project, which is currently being reworked following a lawsuit by Amazon. Schmidt was a member of the Alphabet parent Google until 2019, and has since made investments in Pentagon contractor Rebellion Defense.

The NSCAI concluded its three-year mission on October 1 and was officially closed. Critics fear that the legacy of the NSCAI will be lost, but fans say they are proud of it. Both points to the fact that some of the recommendations from the group, which steered the Pentagon to more closely work with the tech sector, were already incorporated into law. There are few laws in the US that specifically address AI. The commission helped to shape a large portion of them.

NSCAI claims that 19 of its suggestions to Congress were included within the defense budget approved December 2020. One directs that the Pentagon use an industry exchange program to attract more AI talent from tech firms. Another promoted the Pentagon's Joint AI Center director, which seeks to increase military use of AI by taping commercial AI providers like Google to report directly to the deputy Secretary of Defense.

The group also recommends that the Pentagon establish an internal platform to support current and future AI projects. It will use computing and storage services from a pool vetted cloud companies. Another recommendation is to encourage departmental use of commercial AI solutions to automate many administrative processes.

The commission presented a concrete and strategic plan for US technology policy. Martijn Rasser, director of technology, Center for a New American Security.

The commission recommended new investments in AI training, research within and outside of the Pentagon, as well as support for US semiconductor production. What was the overall message of the group? Artificial intelligence is key to the country's safety and destiny. It also helps to compete with China's plans to be the dominant in military and commercial AI.

Martijn Rasser (director of the technology program and national security program at the Center for a New American Security) said that the commission presented a very concrete plan for US technology policy. It is amazing to see so many ideas made it into legislation.

Meredith Whittaker is the faculty director at AI Now Institute at New York University. She has a different perspective on how the commissions work. She says that I saw a highly conflicted quasi-government body, writing legislation and policy from the sidelines. This could potentially bring in hundreds of millions of money for big tech. Whittaker was previously employed by Google where she organized protests against a Pentagon project that used its technology to analyze drone surveillance footage.

A 2018 law created the NSCAI. It had a charter with an annual budget of $5million and 26 employees. Twelve commissioners were appointed by members of Congress, two from the Pentagon, and one from the Department of Commerce.