Unions shape AI policy in workplace, Rutgers study finds
Labor unions can successfully push back against AI expansion in the workplace, according to research from Rutgers University that challenges the assumption that widespread AI adoption is inevitable.
The study examined the 2023 Screen Actors Guild-American Federation of Television and Radio Artists (SAG-AFTRA) strike and contract negotiations, focusing on how unions addressed generative AI and LLM issues. Researchers analyzed union communications, official policies, and media coverage to understand how worker organizations countered what the technology industry portrays as inevitable AI integration.
The researchers created a "spectrum of engagement" to map how unions respond to the narrative that AI expansion cannot be stopped. One end includes engagement strategies: collaboration with tech companies, consent agreements, compensation structures, and policy frameworks. The other end includes disengagement strategies: outright bans, worker-led organizing, and commons-based alternatives.
What the contract achieved
The SAG-AFTRA contract secured protections on consent, compensation, and pay discrimination related to AI use. However, the study notes that AI-related negotiations remain difficult because few laws govern AI in the workplace and widespread belief persists that AI expansion is unavoidable.
To overcome these barriers, researchers recommend defining AI broadly as a subject open to collective bargaining. This ensures workers retain autonomy over technological decisions affecting their jobs.
Global implications
The study emphasizes that AI systems often exploit workers in the Global South through supply chains and outsourced labor. Researchers call for building solidarity across North and South to address these disparities.
The research suggests unions have concrete tools to shape AI policy: collective organizing, explicit refusal, and demands for stronger oversight. AI for Science & Research professionals can apply these findings to understand how institutional and policy-level decisions about technology adoption actually get made.
"AI is not inevitable, and we have the power to come together and refuse it," the researchers concluded.
Your membership also unlocks:
