Google DeepMind employees are collaborating on AI armies

Employees at Google’s DeepMind headquarters have voted to join forces to prevent the company’s AI technology from being used by Israel and the US military. In a letter to Google management on Tuesday, workers called for the Communication Workers Union (CWU) and Unite the Union to be recognized as joint representatives, with 98 percent of CWU members at DeepMind voting to support the move.

“We don’t want our AI models to be against international law, but they are already helping to kill Palestinians in Israel,” DeepMind said in a statement shared with CWU. “Even if our work is used for official purposes, as the administration has repeatedly told us, it is still helping to make genocide cheaper, faster, and more efficient. This must stop immediately, as it must harm the people of Iran and the lives of people everywhere.”

If successful, the contract bid would provide representation for at least 1,000 employees from Google’s DeepMind headquarters in London. Regulators now have 10 business days to voluntarily identify unionization efforts before they are legally required to enforce approval.

A contract bid includes specific needs what employees want from Google, which includes a clear commitment that they will not pursue the development of tools, technologies or agreements that harm or monitor people; discussions about the use of AI that “significantly affect our responsibilities, workload, or job security;” and the right for workers to avoid jobs that conflict with their “moral values.” DeepMind employees around the world are said to be considering public demonstrations and “research boycotts,” where they avoid processing Google AI services such as Gemini AI, as part of a campaign against Google’s military and industrial AI contracts.

We’ve reached out to Google for the answer.

This comes a week after hundreds of Google employees signed an open letter CEO Sundar Pichai wants the company to refuse to sign AI contracts with the Pentagon. Soon after, Google – along with OpenAI and Nvidia – signed contracts which they say allows the US Department of Defense to use their AI models for “any legitimate government purpose.” In 2024, the company they laid off more than 50 workers in response to protests against Google’s military partnership with the Israeli government.

“This is an important time when technology workers at the Google AI lab are connecting with some of the world’s most marginalized people in meaningful ways, based on the basic principles of cooperation and commercial cooperation,” said John Chadfield, CWU’s technology services officer. “By exercising their right to join forces, they have the opportunity to tell their employers to stop disrupting the collective bargaining agreement, which is in line with the demands of many workers in the UK and elsewhere.”