Tech workers once had little reason to organize. High pay, flexible hours, and a genuine sense that they were building the future made unions feel unnecessary. That calculation has shifted. The job market that once let engineers leave on Friday and start somewhere better on Monday no longer exists the way it did.
The numbers tell a clear story. Close to 400,000 tech workers have been laid off since 2025, with more than 150,000 gone this year alone. Several of those departures were explicitly tied to AI investment. Amazon cut roughly 30,000 corporate roles across late 2025 and early 2026. Salesforce shrank its customer support team from 9,000 to 5,000, with leadership stating that AI agents had taken on the work. Accenture cut around 11,000 roles and signaled that workers who could not adapt to AI tools would be let go. This is not confined to tech anymore. Law firms, pharmaceutical companies, and consulting groups are all making similar moves.
The workers who remain are not simply relieved. Many are being asked to absorb the responsibilities of vacant positions while managers expect AI to fill the gaps. A survey of University of California IT staff found that 65% had taken on more work due to unfilled roles, yet only 22% felt secure in their jobs. That gap between expectation and reality is precisely what is driving organizing.
The organizing itself is producing real results. The Alphabet Workers Union delivered a petition signed by more than 4,500 Google employees to senior executives, demanding guaranteed severance and voluntary buyout offers before any involuntary cuts. After that sustained pressure, voluntary exit packages were offered to more than 60,000 Google staff, giving workers more control over their departure than a surprise layoff would. That is a concrete policy change won through collective action at a company with no formal union contract.
Where formal contracts do exist, the protections are more durable. ZeniMax finalized a contract in 2025 that spelled out severance terms, AI usage limits, and overtime rules. When Microsoft cut jobs at ZeniMax shortly after, those protections were binding. Kickstarter's union negotiated a four-day workweek, a minimum salary floor, and a clause preventing management from replacing staff with AI outright. These are not aspirational statements. They are enforceable terms.
The legal environment in the United States complicates the picture. The National Labor Relations Board, the federal body that oversees union elections and worker protections, was left without a working quorum for most of 2025 after the Trump administration fired a board member. Union elections overseen by the NLRB fell by 30% that year, and 59,000 fewer workers participated in those elections compared to 2024. Companies now have more room to push back on organizing efforts without facing immediate legal consequences. Workers in the UK and Europe face a different, often more favorable legal environment, which is part of why Google DeepMind and Meta organizing efforts have moved faster there.
The pattern from tech tends to arrive elsewhere with a delay. Nurses, Hollywood writers, and port workers have already fought AI-related battles in their own contract negotiations. The International Longshoremen's Association negotiated a full ban on automated technology at the ports it covers. The Las Vegas Culinary Workers union secured the right to bargain before any AI system is deployed in its workplaces. Both agreements show the range of outcomes possible: outright prohibition on one end, a seat at the table before decisions are made on the other.
For managers and directors in industries not yet touched by these conversations, the signal is worth reading now. Workers are watching what happens in tech. The contracts being written today will become the reference point when organizing reaches insurance back offices, procurement teams, and logistics operations. The question is not whether workers will want a say in.