Negotiations between Google DeepMind and its London-based employees over unionization began this week but quickly faced obstacles. Union representatives expressed frustration after an initial meeting lacked involvement from DeepMind leadership, signaling a lack of good faith, according to John Chadfield, a Communication Workers Union officer. The meeting, which included union officers, DeepMind HR representatives, and a third-party arbitrator, was described as a time-wasting exercise by Chadfield, who claimed the absence of senior management was a clear indicator of disengagement. The talks were initiated in May when DeepMind employees requested recognition from two unions, but the company denied the request and opted for third-party arbitration instead. The meeting also saw DeepMind HR representatives interrupting an employee reading a prepared letter from colleagues supporting unionization, according to multiple sources.
The letter accused Google of stifling open dialogue by shutting down internal chat venues and reprimanding employees who attempted to discuss unionization outside of formal channels. A DeepMind employee involved in drafting the letter described the actions as union-busting techniques intended to intimidate staff. Google DeepMind spokesperson Al Verney stated that the first step was to define the unions' representatives and agreed on next steps, emphasizing that appropriate representatives attended the meeting. The push for unionization at DeepMind began in February 2025, following Alphabet's removal of a pledge against using AI for weapons development and surveillance. Employees have raised concerns about the militarization of AI models, with staff at DeepMind and OpenAI signing an open letter in support of Anthropic after the US Department of Defense sought to label the lab a supply chain risk.
In April, The New York Times reported that Google had entered a deal allowing the Pentagon to use its AI for any lawful government purpose, with roughly 600 US-based employees protesting the terms. The US Department of Defense confirmed it had reached agreements with seven leading AI companies, including Google, to use their models on classified networks. Google has defended its government deals, stating it remains committed to preventing AI use for domestic mass surveillance or autonomous weaponry without human oversight. In 2021, US Google employees formed the Alphabet Workers Union, which is not recognized by Alphabet for collective bargaining but has successfully negotiated agreements on behalf of contractors. If negotiations in London do not progress, Chadfield said the CWU would seek an arbitration committee to force Google to recognize the unions.
He emphasized that both sides must be willing to make concessions, noting Google has not offered any so far. Google DeepMind spokesperson Verney reiterated that the company would continue to engage constructively with employees and offer other channels for discussion.
Source: wired
