Google to End Involvement in Pentagon AI Project
Google is reportedly abandoning plans to help the Pentagon develop a controversial AI system to analyze footage taken from aerial drones.
Due to employee backlash, the company will not pursue another contract for the Pentagon's Project Maven, Google Cloud CEO Diane Greene told employees on Friday, according to Gizmodo, citing anonymous sources.
Word of Google's involvement in the research prompted thousands of employees to sign an internal letter in protest; a dozen reportedly went so far to resign.
So far, Google hasn't commented on Gizmodo's report. But publicly, the company has defended its work with the Pentagon and claimed the research is focused on "non-offensive" purposes.
"The technology is used to flag images for human review and is intended to save lives and save people from having to do highly tedious work," the company said back in April.
However, Google employees and outside AI experts are concerned that the U.S. military will one day weaponize the research to conduct warfare. "We are then just a short step away from authorizing autonomous drones to kill automatically, without human supervision or meaningful human control," reads an open letter from academics protesting Google's involvement in Project Maven.
According to Gizmodo, Google's current contract for Project Maven lasts until 2019. It was initially valued at $15 million, but over time the budget was estimated to reach as high as $250 million.
Google may have also had plans for Project Maven that went beyond analyzing drone footage. According to Gizmodo: "Google intended to build a 'Google Earth-like' surveillance system that would allow Pentagon analysts to 'click on a building and see everything associated with it' and build graphs of objects like vehicles, people, land features, and large crowds for 'the entire city.'"