Negotiations between Google DeepMind and its London-based employees regarding potential unionization recently hit a significant snag. Initial talks have left union representatives feeling their time was largely wasted, casting a shadow over future discussions. This development follows a request in May by DeepMind employees for Google to officially recognize the Communication Workers Union (CWU) and Unite the Union as their joint representatives.
Google had previously denied this request, though it did agree to engage in negotiations facilitated by a third-party arbitrator. However, the first meeting held on Wednesday proved deeply frustrating for those advocating for unionization. Key union officers and DeepMind employees involved in the push were met only by the arbitrator and DeepMind HR representatives, with no senior leadership present.
Union Frustration and Allegations
John Chadfield, a CWU officer who attended the meeting, voiced strong disappointment, stating that the absence of senior management from the outset signals a lack of good faith. He claimed this is often a precursor to a “time-wasting exercise,” suggesting negotiations have stalled very early on. This sentiment highlights a significant disconnect in expectations for the talks.
Google DeepMind, however, refutes claims that negotiations have stalled. According to Al Verney, a Google DeepMind spokesperson, “The first step in the process is to define who the unions want to represent and the parties agreed on next steps to do this.” Verney also maintained that “The appropriate representatives attended this initial meeting,” indicating a different perspective on the meeting’s protocol.
During the meeting, a DeepMind employee read a letter on behalf of colleagues supporting unionization, which accused Google of treating its workers “as a problem handed off to HR” rather than engaging in meaningful dialogue. Multiple sources reported that DeepMind HR representatives interrupted the employee reading this statement on two occasions. This perceived interference further fueled the union’s concerns about the company’s approach.
The letter went on to detail alleged attempts by Google to suppress open dialogue and dissent among DeepMind employees. These tactics reportedly included shutting down or reconfiguring internal chat forums and preventing staff from responding to company-wide communications about the unionization bid. Employees who tried to circumvent these restrictions were allegedly “reprimanded” by HR, leading to accusations of intimidation.
An anonymous DeepMind employee involved in drafting the letter characterized these actions as “well-established union-busting techniques.” While Google’s Verney reiterated the company’s commitment to “engage constructively” in the process and have “open dialogue with employees,” these allegations paint a picture of a strained workplace environment. Google stated it continues to offer various other channels for employees to discuss their views outside of the formal union process.
The Roots of the Unionization Push: Ethical AI Concerns
The drive to unionize at DeepMind began earlier this year, sparked by a significant policy change at Google’s parent company, Alphabet. Alphabet removed a crucial pledge from its ethics guidelines that committed not to use AI for purposes like weapons development and surveillance. This policy reversal deeply impacted many employees.
One DeepMind employee, who requested anonymity, shared that these ethical principles were a primary reason they joined the company. “We basically just got rid of them all,” they lamented, reflecting a common sentiment among colleagues. This shift fundamentally altered the company’s perceived ethical stance on AI development, prompting many to seek collective representation.
Concerns over the militarization of AI models extend across the industry. In late February, staff at DeepMind and OpenAI demonstrated solidarity by signing an open letter supporting Anthropic. This came after the US Department of Defense sought to classify Anthropic as a supply chain risk due to its refusal to permit its technology for autonomous weapons and mass surveillance.
Adding to these concerns, The New York Times reported in April that Google had entered a deal allowing the Pentagon to use its AI for “any lawful government purpose.” This broad agreement reportedly led to roughly 600 US-based Google employees signing a protest letter. The US Department of Defense later confirmed similar deals with several leading AI companies:
- SpaceX
- OpenAI
- Microsoft
These agreements allow their models to be used on classified networks, intensifying the debate over the ethical implications of AI development and military applications.
Google has consistently defended its collaborations with government organizations. Jenn Crider, a Google spokeswoman, previously stated, “We are proud to be part of a broad consortium of leading AI labs and technology and cloud companies providing AI services and infrastructure in support of national security.” She emphasized Google’s commitment to the consensus that AI should not be used for domestic mass surveillance or autonomous weaponry without appropriate human oversight.
What Lies Ahead for DeepMind Unionization
This isn’t Google’s first encounter with employee unionization efforts. In 2021, US-based Google employees formed the Alphabet Workers Union. While this union is not recognized by Alphabet for collective bargaining, it has successfully negotiated agreements on behalf of Google contractors, setting a precedent for organized labor within the company.
Should the current negotiations in London fail to progress, the CWU representative, John Chadfield, indicated that employees would escalate the matter. They plan to ask an arbitration committee to compel Google to formally recognize the unions. This move would significantly intensify the pressure on the tech giant.
“We’re hoping that Google genuinely comes to the table and we can agree something amicably,” Chadfield expressed, highlighting the union’s desire for a constructive resolution. However, he stressed the need for mutual concessions, claiming, “Google is coming with no concessions whatsoever,” suggesting a challenging road ahead for DeepMind employees seeking collective representation.
Source: Wired – AI
