What Unions Should Ask Before AI Shows Up at Work
AI rarely arrives at a workplace with an announcement. It shows up as a new feature in the scheduling software, a "productivity" dashboard, a screening tool in hiring, or a chatbot that starts handling work people used to do.
By the time members notice, decisions have often already been made. That is why the most useful thing a union can do is ask good questions early, and keep asking them.
This piece is about two things: what to ask employers about AI, and how unions can use AI responsibly in their own work. It is practical guidance, not legal advice. Your counsel should weigh in on anything that touches bargaining obligations or contract language.
Part one: Questions to put to the employer
You don't need to be a technologist to ask these. You need clear, plain answers in writing.
What is being used, and why?
- What AI or automated tools are in use now, or planned, that affect bargaining unit work?
- What problem is each tool meant to solve?
- Who is the vendor, and what does the tool actually do day to day?
What does it decide, and who checks it?
- Does the tool make or recommend decisions about hiring, scheduling, assignments, pay, discipline, or performance?
- Is a person required to review those recommendations before they take effect?
- Can a worker find out when AI played a role in a decision about them, and challenge it?
What data is collected about workers?
- What information does the tool collect: keystrokes, location, calls, video, productivity metrics?
- Who can see it, how long is it kept, and can it be used for discipline?
- Is worker data shared with the vendor or used to train its products?
What happens to jobs and skills?
- Will any duties be moved, reduced, or eliminated because of the tool?
- What notice will the union receive before changes take effect?
- Will workers be trained on new tools on paid time?
Many unions are exploring contract language around notice, consultation, human review, data use, and training. What fits depends on your contract, your sector, and the law that applies to you, so work through specific proposals with counsel.
Part two: Using AI inside the union
Unions are often short-staffed and member-funded, which makes saving staff time valuable. Used carefully, AI can help with work that is repetitive and easy to check:
- Member communications. First drafts of newsletters, meeting notices, and plain-language explainers of contract changes, reviewed by staff before sending.
- Translation support. Rough drafts in other languages, checked by a fluent member or staff person before use.
- Research. Summarizing long public documents, such as published reports or employer filings, to speed up preparation.
- Internal organization. Turning meeting notes into action items, or organizing a large set of public materials.
Where to draw firm lines
The same concerns you raise with employers apply to your own use. Some information should stay out of general AI tools entirely:
- Grievance files and anything tied to a specific member's case.
- Bargaining strategy, internal positions, and proposals before they're public.
- Member lists, contact details, and dues records.
- Organizing campaign details and the names of people involved.
Put these rules in a short written policy, name approved tools, require human review of anything going to members, and train staff and officers on it. Practicing what you ask of employers also strengthens your position at the table.
A simple way to start
- Inventory the AI tools you know employers are using, or plan to use.
- Turn the questions above into a standing information request.
- Adopt your own internal AI use policy.
- Train officers and staff so they can spot AI in the workplace and use it well in the union office.
Preparing for AI questions at the table, or inside your own office? We help unions understand the technology in plain terms and put sensible rules in place. We are independent and don't sell software. Schedule a call →