Seeking Worker Input in the Use of AI

Worker voice belongs at the center of how AI gets implemented. Partnership on AI developed four case studies on that premise, examining how employers and unions are responding to AI tools in real workplaces. Rather than debating AI’s theoretical potential to upend our economic lives, we looked at how those conversations are actually unfolding today.

These case studies draw on interviews with representatives from two employers that deploy AI in the workplace — one private sector, one public sector — and three labor unions that have secured AI-related job protections for some 30,000 workers in the United States, Italy, and Ireland. Though they span multiple industries and very different labor markets, the cases point to a clear set of insights for leaders in both labor and management.

Ultimately, we found that employers and labor leaders alike benefited from talking openly about the use of AI. When workers shared their expertise, management gained a better understanding of shared goals and the circumstances that make AI deployment successful.

Shared insights for labor and business leaders:

• Business and labor leaders can build on workers’ expertise to identify opportunities and risks in AI and to craft higher-order policies broad enough to address rapidly evolving technologies.
• Defining AI as a tool, not a replacement, establishes a foundation for addressing risks of worker displacement while enabling organizational innovation.
• Build on established worker-employer trust to make AI negotiations more productive — and even use AI negotiations to deepen that trust.
• Outside perspectives can help both business and labor leaders think more creatively about how to approach AI implementation and governance.

Key insights for business leaders:

• Responsible AI teams can support HR teams in engaging workers on AI policy by using their expertise to uncover gaps between AI governance guidelines and workplace practices.
• Employers establish more effective AI guidelines when they treat workers as experts and find multiple channels for gathering feedback from their employees.

Key insights for labor leaders:

• Unions can build on existing labor laws and previous agreements regarding technology in the workplace despite the seeming newness of AI tools.
• Labor leaders may need external support to design adequate long-term protections against potential job displacement.