Google DeepMind: Employees demand exit from AI military contracts
Almost 200 employees of Google DeepMind are concerned that the AI technology they have developed is being used for warfare.
(Image: testing/Shutterstock.com)
Within Google DeepMind, the British subsidiary of the US internet company that specializes in artificial intelligence (AI), there is growing discontent over the support of big data solutions for military conflicts. Almost 200 DeepMind employees – around 5 percent of the workforce – recently signed a letter calling on the tech giant to terminate its contracts with military organizations. This was reported by the US magazine Time. The authors of the letter dated May 16 did not name any specific armed forces, as they were not concerned with a specific conflict. However, they refer to reports that Google has a direct contract to provide cloud computing and AI services for the Israeli military (IDF) and that – is part of the relevant Nimbus project together with Amazon. This is to be used in the Gaza war for mass surveillance and target selection.
"Any involvement in the military and weapons production compromises our position as a leader in ethical and responsible AI and violates our mission and stated AI principles," reads the letter, which is circulating within DeepMind, according to Time. The signatories are demanding that DeepMind's managers block military users' access to in-house technologies. They are campaigning for a new supervisory body to ensure compliance with this approach.
Great frustration within the workforce
Three months after the letter was circulated, Google has not responded to any of the appeals, according to those involved. "We have not received any meaningful response from the company management," one of the employees complained to Time. "We are getting more and more frustrated". When Google bought DeepMind in 2014, the heads of the lab at the time promised the company that their AI technology would never be used for military or surveillance purposes. According to the report, a Google spokesperson has now affirmed that the company is adhering to its AI principles. The Nimbus contract only applies to work "performed by Israeli ministries in our commercial cloud that agree to abide by our Terms of Service and Acceptable Use Policy". However, the guidelines do not rule out using Google AI to serve military customers.
Videos by heise
The Guardian wrote in April that 50 Google employees who were critical of the cooperation with Israel's military had already been dismissed in the first months of 2024. According to military insiders, a veritable "kill cloud" has long since connected sensors and weapons platforms for drones with a globally distributed network of devices, software and a multitude of other nodes via satellites, cables, radio and digital communication links. Cloud-based solutions, including those from major US providers, are thus "used for analysis, targeted strikes and killing".
(nie)