Google has confirmed that its technology is being used by the Pentagon to analyse footage captured by drones.
A program called Project Maven is utilising the technology to automate the analysis of objects in the enormous amount of images that are captured by the Department of Defence’s surveillance drones – also known as unmanned aerial vehicles (UAVs).
Gizmodo reported that some Google employees were “outraged that the company would offer resources to the military for surveillance technology involved in drone operations”.
There have been almost 30,000 coalition strikes against targets in Iraq and Syria since the US-led intervention in 2014, the intelligence behind many of which is developed by analysis of UAV surveillance footage.
Google confirmed that software called TensorFlow was being used by the Pentagon and said: “The technology flags images for human review, and is for non-offensive uses only. Military use of machine learning naturally raises valid concerns.
“We’re actively discussing this important topic internally and with others as we continue to develop policies and safeguards around the development and use of our machine learning technologies.”
The open-source TensorFlow software is widely used in machine learning applications which attempt to allow computers to be taught how to identify objects in video footage.
The algorithm is identifying cars and trees and it is designed to assist the limited number of human analysts detecting threatening objects in an increasingly unmanageable amount of footage.
Project Maven “is a pilot with the Department of Defence, to provide open-source TensorFlow APIs that can assist in object recognition on unclassified data”, a Google spokesperson told Gizmodo.
The DoD said that Project Maven was initially focusing on identifying 38 classes of objects which were important for its analysts to detect as the coalition continues to target Islamic State forces.
The project’s head, Marine Corp Colonel Drew Cukor, told an audience of military and technology experts that the defence sector needed to embrace artificial intelligence technologies such as machine learning, according to Gizmodo.
“There is no ‘black box’ that delivers the AI system the government needs, at least not now,” Colonel Cukor said.
“Key elements have to be put together… and the only way to do that is with commercial partners alongside us.”