The potential for artificial intelligence (AI) and the push toward more autonomous uses of technology have already seen success in the commercial space. It’s no surprise then that military application of these technologies is a huge priority for the U.S. Department of Defense (DoD). It has recently come to light that the DoD has tapped into Google’s artificial intelligence prowess as part of an endeavor to keep military tech at the cutting edge. The Pentagon has expressed concerns that Russian and Chinese innovation may be outpacing the US in the areas of AI and Quantum Computing.
In April 2017, in order to address this concern, the DoD established the Algorithmic Warfare Cross-Functional Team (AWCFT). Also known as “Project Maven“, the project looks to close this potential gap by significantly speeding up the integration of big data, AI, and machine learning into military operations, an endeavor the DoD isn’t wasting any time pursuing. The DoD has already announced active development of drones capable of choosing targets without human involvement in the decision process.
According to Google, their specific involvement in Project Maven entails the DoD’s use of Google’s TensorFlow open source software library. This software serves as the basis of the AI tech that will allow the DoD to quickly sort through hundreds of thousands of still and moving surveillance footage in order to accurately identify objects of interest, faces and even behavioral patterns. Referred to as “Computer Vision”, the DoD is essentially trying to teach drones to understand visual information similar to the way the human mind does.
While Google has a history of being very selective about participation in defense-related contracts, more recently, this has seemed to change. Current Google exec, Milo Medin and former Alphabet executive chairman, Eric Schmidt both currently serve as members of an independent federal committee know as the Defense Innovation Board. This committee actively advises the Pentagon on the use of cutting-edge tech in defense strategies including, but not limited to the collection, use, and analysis of data.
Google states that the tech only flags images for human review and is not being used in combat situations. The issue has become a heated topic raising concerns especially for some Google employees who disagree with the company’s involvement with military contracts. So much so that more than 3,000 Google employees have signed a petition in protest and some have even resigned their position, urging the company to pull out of military projects. More recently it has been reported that because of the negative feedback, Google will not renew the Project Maven contract.
While innovation in the area of defense is needed, the apprehension that these innovations can be abused is real and must be addressed meaningfully. The potential for improved drone targeting could very well save civilian lives in the long run but the same tech, if not regulated appropriately, could have just the opposite effect.