Hi, I’m Katie Malone, and this is Innovation Wire by Technical.ly. We send this newsletter every two weeks, covering local trends in AI, tech, startups, entrepreneurship and jobs of the future.
If this was forwarded to you, subscribe here.
Today we’re talking about AI and labor union contracts.
The debate over AI in the workplace often shows up in think pieces and panel discussions. In union contracts, it quickly becomes concrete.
Contracts have to answer practical questions: When does management have to tell workers a new system is coming, and how many details do they have to share? Can software be used to track performance? What happens if automation changes a role, speeds processes or shifts responsibility onto fewer people?
This isn’t a new conversation, but AI is causing such rapid shifts that it’s become a challenge to keep up.
“Unions have been negotiating over technological change for decades,” Lisa Kresge told me. She’s a senior researcher in the Technology and Work Program at the UC Berkeley Labor Center. “But today’s agreements are becoming more specific and more directly tied to how particular systems are implemented and used on the job.”
Kresge compiled an inventory of related provisions in more than 500 union contracts. Many agreements still refer to “technology” in general terms, but there’s been a noticeable shift toward writing rules for certain tools, including generative AI and algorithmic decisionmaking.
Provisions now often focus on process, she said. Workers want notice before new systems are introduced, access to information about how they work and a formal role in discussions about how they are used.
Contract language also addresses what happens after the technology arrives: retraining, job protections, limits on surveillance or requirements that human oversight remain in decisions involving safety or discipline. Some include guarantees that employees can’t be blamed for errors caused by automated systems, or that productivity gains cannot lead to heavier workloads.
It’s not just happening in industries that have drawn the most attention, such as film, television and news. This kind of language is also showing up in union contracts in healthcare, manufacturing, transportation, education and government jobs, Kresge said.
Only about 1 in 10 US workers belongs to a union, and union jobs are concentrated in certain sectors. The guardrails written into one contract may reflect risks that don’t exist elsewhere, said Jeremy Schwartz, an economist at Loyola University Maryland. Teachers may see AI as a tool that makes lesson planning easier, for example, while longshore workers, who have spent years fighting automation, tend to view new systems as a direct threat to their jobs.
Schwartz also noted that companies are under pressure to make their AI investments pay off, often by trying to do more with fewer workers.
With the federal stance of nonregulation, and the highly varied approaches at the state and local level, union contracts are one of the few mechanisms that ensure AI adoption comes with terms attached.
What gets written in now will shape the balance between caution and speed for years to come.
Another angle:
“The AFL-CIO’s efforts to limit AI-driven productivity will only lead to slower wage and economic growth. Rather than engaging in a Luddite-inspired campaign against a critical new technology, organized labor should push Congress to improve programs for displaced workers.”
How has AI impacted your job or management style? Hit reply and let me know.
Technical.ly headlines of the week
3 things to know
🏀 March Madness: Getting your entire bracket entirely right may be statistically impossible, but this analytics professor in Georgia says he’s developed an algorithm that’ll get him pretty close — and says he’s doing it better than any LLM. [Georgia Tech]
🧑🌾 Harvesting data: In Vermont, generations of oral history help guide farming decisions. Researchers say AI could preserve and scale that knowledge, but some farmers are skeptical. [WCAX]
🚦 Red light, green light: Automated license plate readers are spreading nationwide, but in Las Vegas, reporting finds police are avoiding public scrutiny by funding the systems outside the normal taxpayer process. [Nevada Independent]