For decades, tech workers were the people least likely to join a union. That’s changing, and AI is a big part of why. A widely discussed report on Hacker News describes engineers at Google DeepMind, Meta, Alphabet and the University of California organizing because of AI-linked layoffs, heavier workloads, workplace surveillance and arguments over how AI products get used.
The irony is hard to miss. The people building AI are among the first to feel its downsides at work.
🏴 What’s changing
The old deal in tech was simple: high pay, generous perks and a flat structure where engineers felt like equals with their bosses. Not many people saw a reason to organize. Now the numbers are moving:
- Alphabet Workers Union (AWU) represents more than 1,400 workers across Google, YouTube and Waymo. Total union actions at Alphabet doubled in 2025 compared with the year before.
- AWU’s job security petition got 4,500 signatures and was delivered to executives last week.
- Google DeepMind and Meta staff in the UK are trying to unionize. Part of their objection is about AI being used for military purposes and to monitor employee productivity. DeepMind says it’s in the “early stages” of recognizing the union and plans to “continue to engage in good faith.” Meta didn’t comment.
- University of California IT staff reported in a survey that 65% had taken on extra work because of a vacant position. Only 22% said they felt secure in their jobs.
💼 Why the fortress cracked
Job security used to be something tech workers took for granted. John Chadfield, co-founder of United Tech and Allied Workers, the UK’s only union just for tech workers, remembers when you could leave a job and land a better-paid one within a week. His message to engineers now: “You might be well paid, but you are still workers.”
Simone Robutti of the Tech Workers Coalition is even blunter: “It’s clear that with very few exceptions, tech workers are as replaceable as other workers.”
Then there’s the productivity problem. AI was supposed to make work lighter. Max Belasco, a business systems analyst at UCLA School of Law, says managers expect AI to fill gaps left by vacant roles, and people just end up doing more work.
⚖️ Two sides of the table
On the worker side, a Meta employee involved in the UK drive said a union “is the best way to actually show up as owners, and take our seat at the table.”
Employers describe it differently. A UC spokesperson said the system is using AI “thoughtfully and responsibly to strengthen its operations,” and that its choices follow operational needs plus privacy, security and policy concerns.
Organizational psychologist Katie Hawkes points to the gap between those two views: “when employees begin organizing, it’s often a signal that they don’t feel heard through the channels that already exist.”
🔭 The next 1-3 years
AWU predicts AI will drive more organizing. Keep in mind that a union has every reason to say that. Still, its own numbers back it up, since actions doubled in a single year. Here’s what I expect:
- AI clauses in contracts become normal. SAG-AFTRA already fought over AI replacing creative workers, and National Nurses United has pushed back on AI in clinical settings. Tech unions will go after similar terms on surveillance, layoffs and how products get used.
- Tech becomes the test case. How Silicon Valley workers do is likely to shape what happens as AI moves into finance, healthcare and the public sector.
- Ethics turns into a bargaining issue. Disputes over military contracts and monitoring tools used to stay in internal forums. Now they’re moving into formal negotiations.
✅ What businesses should do now
- Measure real workload after AI rollouts. Don’t assume the savings showed up. Ask the people doing the work.
- Be open about monitoring. Tracking productivity in secret is a fast way to start a union drive.
- Build feedback channels that actually work. If people feel heard on how AI gets deployed, they’ve got less reason to organize around it.
- Plan for AI terms in employment talks. Whether or not a union shows up, workers will expect a say.
AI was supposed to make tech workers more powerful. Instead, it’s reminding a lot of them that they’re employees. Expect the companies building AI to be some of the first to negotiate over it. You can find the full story at the original source.