Software engineers at Meta and Amazon find themselves in an unusual spot this spring. Layoffs continue to sweep through their ranks. AI systems trained on their own code and keystrokes threaten to shrink teams further. And public sentiment appears to favor workers who push back.
The signs point one way. A fresh piece in Gizmodo captured the mood. Published today, it argues developers should consider banding together before AI-driven cuts accelerate. The piece cites tens of thousands of tech job losses already this year. Companies redirect billions into artificial intelligence infrastructure. Human roles shrink in response.
But the story runs deeper. At Meta, employees recently distributed flyers across U.S. offices. They protested new software that records mouse movements and keystrokes. The goal? Gather data to train AI models. Reuters reported the action on May 12, describing it as the most visible sign yet of a budding labor effort inside the social media giant. Some UK-based Meta staff have gone further. They launched an organizing drive with United Tech and Allied Workers, a branch of the Communication Workers Union. Their website plays on a famous book title. It reads Leanin.uk.
And. The anger feels real. Workers see their daily inputs feeding systems meant to replace parts of their jobs. One Meta contractor in Ireland, Covalen, plans to lay off 700 content moderators. Many trained AI systems now edging them out. UNI Global Union called on Meta to take responsibility for these supply-chain workers. The group demanded notice before AI introductions, training tied to continued employment, and the right to refuse training one’s own replacement. Christy Hoffman, UNI Global Union’s general secretary, stated the case plainly. Tech companies treat the workers whose labor and data built AI as disposable.
Amazon presents a parallel picture, though its battles have played out more among warehouse staff than coders. The company spent more than $26 million on anti-union consultants in 2025 alone. Yahoo Finance detailed the figure in April. That sum set a record. It covered payments to at least six specialized firms. Average consultant fees ran above the typical annual wage of an Amazon warehouse worker. Federal filings captured only part of the picture. Legal costs for fighting organizing drives often add more.
Despite the spending, progress occurs. The National Labor Relations Board ruled in April that Amazon must recognize and bargain with the Teamsters at its Staten Island JFK8 facility. The Teamsters hailed it as historic. Four years after workers first voted to unionize, the e-commerce giant received its first formal order to negotiate. Similar efforts spread to other sites. Organizers in North Carolina continue challenging election results at a Garner facility. They allege coercion and intimidation. LaborLab tracked Amazon’s payments to a union-busting consultant at $2,200 per day at one Durham location.
Developers watch these warehouse fights closely. Many sense the same forces at work in engineering offices. Rumors of heavy layoffs targeting senior engineers circulate at Amazon. AI metrics drive strange behaviors. Some employees practice malicious compliance. They over-automate tasks to boost numbers that justify further automation. The pattern echoes across the industry. LinkedIn reportedly cut 5 percent of its workforce recently. Meta’s own chief executive, Mark Zuckerberg, told staff that AI costs contributed to the elimination of 8,000 positions.
Public opinion adds pressure. An AFL-CIO poll of more than 1,500 Americans, referenced in the Gizmodo report, delivered striking numbers. Ninety percent want a human as the final decision maker on AI issues that affect workers. Large majorities back guardrails against harmful workplace AI and demand transparency from employers. Trust levels tell another tale. Only 17 percent trust Democratic lawmakers on these protections. Ten percent trust Republicans. Six percent trust their employers. Nearly two in five trust labor unions. That figure more than doubles support for any other institution.
Such numbers matter. They suggest a broader shift. Tech workers once viewed unions as unnecessary in a high-compensation field. Stock grants and rapid promotion tracks dulled the appeal. Those assumptions erode now. Job security feels fragile. Performance reviews incorporate AI productivity scores. Surveillance tools track more than ever. Engineers who once prized autonomy start asking what collective voice could achieve.
But. Organizing software developers differs from warehouse campaigns. Tech pay scales create divisions. Remote work scatters potential members. Non-compete agreements and strict confidentiality rules complicate coordination. Companies deploy sophisticated tactics to discourage union talk. Still, small groups test the waters. Alphabet Workers Union continues its efforts at Google despite setbacks. Meta’s UK drive and U.S. flyer campaign signal similar stirrings.
The AI transition sharpens the debate. Billions pour into models that learn from employee output. Those same models then optimize teams downward. Hoffman captured the frustration. Experience and human judgment disappear when companies discard workers who built the systems. Developers possess unique insight. They understand how these tools work. They see where automation falls short. A union could press for transparency in AI deployment, limits on surveillance, and real input on how new technology reshapes roles.
Recent contractor layoffs at Meta’s Irish operations illustrate the risk. Over 700 jobs gone in one round. Another 400 earlier. A six-month cooldown prevents many from moving to competing Meta vendors. The Communications Workers Union fought for better severance and called strikes. Success remains partial. The pattern repeats. Companies outsource the human labor that trains AI, then cut those same workers once models improve.
Amazon’s record anti-union spending reveals corporate priorities. Twenty-six million dollars buys a lot of persuasion. It funds town halls, consultants, and monitoring. Yet warehouse victories at Staten Island and growing Teamsters presence among delivery drivers show limits to that approach. An NLRB settlement in March forced Amazon to restore unpaid time off docked from striking workers. The company also agreed to post notices affirming organizing rights.
So the ground shifts. Not quickly. Not without friction. Engineers at Meta who circulated petitions against tracking software took risks. UK organizers built a public website. These steps break years of silence. They coincide with warehouse workers winning bargaining orders and challenging election interference in North Carolina.
Industry watchers note the contrast. Tech once celebrated its meritocracy and individualism. That culture now collides with concentrated power at a few dominant firms. Four companies absorbed much of the $725 billion spent on AI infrastructure, according to one recent analysis. The same firms trim headcount. Workers connect the dots.
Challenges remain steep. Union elections in tech rarely succeed on first tries. Retribution fears deter many. Legal battles stretch for years. Yet the alternative looks less attractive. Continued layoffs justified by AI investment. Expanding surveillance under the guise of productivity. Decisions about human roles made by algorithms trained on data taken without full consent.
The Gizmodo piece ends on a direct note. With trust in lawmakers and bosses so low, workers should trust each other. Unions offer one proven vehicle for that trust. Whether Meta and Amazon developers act on the moment remains uncertain. The ingredients exist. Widespread dissatisfaction. Concrete examples of invasive AI tools. Record company spending to fight organizing. Strong public support for worker protections in the AI age.
Events of the past week alone reinforce the point. Flyer campaigns at Meta offices. New details on contractor cuts in Ireland. Fresh reporting on Amazon’s consultant bills. A Tech workers union drive gains quiet momentum in Britain. Each development adds weight to the argument. Software engineers may soon decide that going it alone no longer serves them. Collective bargaining could become their most effective code review yet.

Pingback: Tech’s Next Reckoning: Why Software Engineers At Meta And Amazon Face A Pivotal Moment For Collective Action - AWNews