98 percent. That's the share of Google DeepMind headquarters staff who voted in favor of unionizing - a near-unanimous result driven by concerns about how the company's AI is being deployed militarily.
In a letter sent to Google management on Tuesday, workers requested that the Communication Workers Union (CWU) and Unite the Union be jointly recognized as their representatives. The central demand: a formal mechanism for employees to weigh in on whether DeepMind's technology is used by US military or Israeli defense forces.
The Precedent Workers Are Pointing To
This isn't Google's first encounter with employee pushback over defense contracts. In 2018, engineers walked out over Project Maven, a Pentagon contract that used Google AI to analyze military drone footage. The protest worked - Google chose not to renew. The DeepMind workers are making a calculated bet that formal collective bargaining is a more durable lever than one-off protests, especially as AI defense contracts grow in value and visibility.
DeepMind's research portfolio makes it a particularly attractive target for military procurement. Work in reinforcement learning (training AI through trial-and-error to make sequences of decisions), logistics optimization, and scientific prediction has clear dual-use potential beyond its published research applications.
What Comes Next
DeepMind's London headquarters means this plays out under UK labor law, which gives workers a route to statutory recognition even if the employer refuses - provided they can show majority support. A 98% vote is hard to dispute on that front.
Google has not yet publicly responded to the recognition request. The company's answer in the next few weeks will signal how seriously it takes this kind of internal opposition as AI military spending increases across the industry. A fight against recognition would likely draw more attention to the contracts that workers are already objecting to.