In 2026, the fight over artificial intelligence in the U.S. has moved out of tech conferences and into union halls, courtrooms, and suburban town meetings.
Workers are demanding limits on algorithmic management. Artists and publishers are suing over training data. And residents living near sprawling data centers are pressing local officials for hard numbers on power and water use. The common thread: a growing insistence that AI can’t keep expanding on autopilot, especially when the costs land on everyone else.
This isn’t one unified movement. It’s a patchwork of conflicts, jobs, copyright, surveillance, electricity, water, colliding with the breakneck rollout of AI systems by the biggest names in tech.
Hollywood put generative AI on the bargaining table
The most visible flashpoint has been Hollywood, where writers, actors, and crew members have forced AI into the center of labor negotiations.
The Writers Guild of America (WGA) has pushed for guardrails aimed at stopping studios from using AI-generated scripts to shrink writers’ roles, or dodge paying them. Studios argue the tools can speed up early-stage work and cut costs. The unions’ response is blunt: there’s a difference between software that assists and software that replaces.
Actors, represented by SAG-AFTRA, have raised alarms about digital replicas, faces, voices, even full-body scans. The fear is that performers, especially background actors and early-career talent, could be pressured into signing broad permissions that let studios reuse their likeness across multiple projects with little real negotiation.
That Hollywood playbook is now being watched closely by other professions, journalists, illustrators, translators, musicians, who see collective bargaining as one of the few ways to force rules, documentation, and compensation when human work becomes training fuel for machines.
In Virginia, data centers are turning AI into a neighborhood fight
AI isn’t just software. The biggest models run on physical infrastructure, massive server farms that are increasingly landing near residential areas.
In Virginia, especially in the fast-growing northern part of the state, residents say they’re surrounded by hulking data center buildings, new transmission lines, backup generators, and construction traffic. For many locals, the debate isn’t about abstract “AI risk.” It’s about what they can see and hear from their front porch.
Data centers can bring tax revenue and construction jobs, but they also raise immediate questions for county boards and planning commissions: How much electricity will these facilities draw? Can the grid handle it? What about noise, land use, and setbacks from homes?
Water has become another pressure point. Not every facility cools servers the same way, but residents want clear disclosures about how much water is needed, especially during heat waves. Underneath that is a bigger political question: should local resources be committed to infrastructure that trains AI products sold worldwide, while communities still struggle to fund schools, housing, and transportation?
Cloud companies point to renewable energy contracts and efficiency plans. Local groups are increasingly demanding proof, audits, enforceable conditions, and public transparency before permits are approved.
OpenAI faces a courtroom war over training data
The backlash is also playing out in lawsuits, as authors, photographers, illustrators, and news organizations challenge the use of copyrighted material to train AI models.
One of the most closely watched legal battles targets OpenAI, with plaintiffs alleging the company used large bodies of protected work without adequate permission. AI companies often argue that analyzing publicly available content is “transformative”, a key concept in U.S. copyright law that courts are still wrestling with in the AI era.
The New York Times has become a central player because the stakes are existential for publishers: if a chatbot can spit out a detailed summary of an article, readers may never click through, subscribe, or see ads. In that scenario, publishers argue, AI becomes a competitor built from their own archives.
The Authors Guild is pushing a similar argument for books, calling for permission, payment, or licensing systems for training on published works. Developers warn that licensing requirements could splinter access to data and slow innovation. Judges will be asked to balance copyright protections, the freedom to compute over text and images, market competition, and the public interest.
Even before final rulings, behavior is shifting. Some publishers are blocking web crawlers. Others are negotiating licensing deals. Photo agencies, including Getty Images, are tightening protections. A once-radical idea is gaining traction in the digital economy: cultural data isn’t just “out there”, it’s accumulated labor with a price tag.
Amazon, Uber, and the rise of algorithmic bosses
In warehouses, delivery networks, and ride-hailing apps, AI often shows up less as a chatbot and more as an invisible manager, setting quotas, tracking performance, optimizing routes, and triggering discipline.
At Amazon, unions and worker advocates have long criticized software-driven productivity demands and granular monitoring of breaks and movement. The complaint isn’t simply that algorithms exist, it’s that workers can be penalized or lose hours without understanding the rules the system is using.
That information imbalance is a core demand for labor groups: access to the criteria behind automated decisions. The Teamsters, a major U.S. union with deep roots in trucking and logistics, have argued that technologies reshaping work should be negotiated, not imposed.
For Uber and delivery platforms, drivers have focused on how algorithmic decisions hit their paychecks: dynamic pricing, account deactivations, trip assignments, and customer ratings. Driver groups are pushing for more transparency, real appeal processes, and pay floors that don’t swing wildly with constant behind-the-scenes tweaks.
Autonomous vehicles add another layer. Companies sell automation as a solution to labor shortages and rising costs. Workers see a direct threat to jobs, and worry that rushed deployments could put public safety at risk. The result is a widening U.S. fight over who controls AI: the companies building it, or the people living and working under it.
Key Takeaways
- The American pushback against AI brings together unions, artists, local residents, and gig workers.
- Hollywood put generative AI at the center of labor negotiations.
- Data centers are sparking local conflicts over energy, water, and land use.
- Lawsuits against OpenAI raise the question of the value of cultural data.
- Workers are demanding more transparency about algorithmic decisions.
- En 2026, aux États-Unis, syndicats, tribunaux, artistes et riverains de data centers, ce que les géants de l’IA affrontent - juillet 9, 2026
- Poitiers : la CGT appelle les chauffeurs Vitalis à la grève face à la canicule - juillet 8, 2026
- Voitures électriques d’occasion: Avere-France chiffre le T2 2026 à 72 244 ventes - juillet 8, 2026