Google Ends Controversial Pentagon AI Contract Amid Employee Protests
Google announces it will let its Project Maven contract with the Pentagon lapse in 2019 and will not pursue future military deals. The decision follows massive internal backlash and the upcoming release of a new AI ethics policy.
Google officially ends its involvement with Project Maven, a controversial Pentagon program that uses artificial intelligence to enhance object recognition in military drones. Google Cloud head Diane Greene tells employees that the company allows its current Defense Department contract to lapse in 2019 and does not pursue a new one. This move aligns with Google's promise to draft a strict ethics policy that explicitly bans the use of its artificial intelligence in weaponry.
The decision directly results from massive internal and external backlash against the military partnership. Since the news becomes public in March, over 4,000 Google employees sign an open letter demanding the company cancel the contract, while dozens resign in protest. Additionally, executives at DeepMind, Google's London-based AI subsidiary, successfully distance themselves from the project by citing a 2014 acquisition agreement that prohibits the use of their technology in military and surveillance systems.
Internal emails reveal that despite the ethical concerns, Project Maven serves as a stepping stone for Google to secure highly lucrative government contracts. The project expects to generate $15 million in revenue over 18 months, with projections reaching up to $250 million in the coming years. Under the contract, Google provides its TensorFlow AI framework and plans to build a sophisticated, Google Earth-like surveillance system for Defense Department analysts to track buildings, vehicles, and people.