"Should unions adopt AI?" 🤨 That's the WRONG question

I'm more concerned about unions choosing to not even engage with increasing AI use among their staff than I am about whether they should adopt AI

Earlier this month I had the pleasure of speaking with Larry Williams, founder of UnionBase, about the role of unions in the AI conversation.

Our discussion covered everything from practical AI applications for unions to the ethical challenges of engaging with technology that impacts workers' livelihoods.

You can watch the full interview below (taken from UnionBase's YouTube channel).

Why unions need to engage with AI

One of the central questions we explored was whether unions should adopt AI tools. As I explained to Larry, I think this framing is not helpful. The real question of today isn't whether unions should adopt AI, but whether they will engage with it.

âťť

"Should unions decide to adopt AI isn’t the best question... what I worry about is that they don't make a decision to engage at all… not making a decision is a decision."

The fact is, AI technology is being adopted at unprecedented speed across workplaces. This includes in unions. With Unions21, I’ve trained 100s of union staff in the last few months on understanding Responsible AI Fundamentals, but this is just a drop in the ocean compared to the number of untrained union staff already experimenting with AI. I hear more stories and see more examples every week.

How unions are already using AI

When Larry asked about practical applications of AI for unions, I shared findings from our ongoing survey of union staff - which is still available for union staff to complete. The top three areas where the 230+ union staff surveyed are already experimenting with AI are:

  1. Text development (52% had tried) – writing, editing, and improving communications

  2. Transcription and translation – making multilingual communication more accessible

  3. Research (35% had tried) – analysing data and information to support campaigns

We're also seeing innovative applications, like this one in case management, with one union officer reducing the time needed to draft case letters from 30 minutes to just 3 minutes using AI assistance.

Do you work in a union and are you looking to learn more about AI?

With Unions21’s Responsible AI Fundamentals for Union Staff courses you can go beyond the hype around AI, understanding the real opportunities and challenges of AI for you, your team and your union. Register today for our instant start self-paced course and join the only training course focused on helping union staff understand AI for their work. Find out more and start learning now >

Barriers facing unions in the AI landscape

Our conversation highlighted two significant barriers unions face when approaching AI, and these contribute to the lack of engagement:

  1. Leadership and staff feeling overwhelmed: Many union leaders and staff don't know where to start with AI. The technology evolves rapidly, and without proper support and training, the prospect of engaging with it feels daunting.

  2. Staff using AI without guidance: At the other end of the spectrum, we're seeing union staff who are already using AI tools but without proper training or organisational support. This creates potential risks around data security, consistency, and effectiveness.

These challenges underscore the need for structured training programmes for unions, and much deeper engagement from all levels of unions. I’m particularly focused right now on the importance of building more learning and engagement at leadership level. Over coming weeks we’ll be releasing more content showcasing how unions have developed AI policies, and what we’re learning from our new courses on managing AI, design for union leaders (get in touch if you’re a union leader and want to find out more!)

Almost 80% of respondents to our survey of union staff have picked out the importance of training to support them in understanding AI’s impact on their work.

Balancing technology adoption with worker advocacy

One of the most thought-provoking parts of our discussion addressed how unions can both organise workers in tech companies while advocating for those affected by AI-driven changes.

I believe requires a values-driven approach where unions are transparent about how and why they're using AI. As I noted in the interview:

âťť

"AI should be about making your impact greater, not about cutting your costs. And that's one of the biggest problems with the whole big tech agenda right now... the way that it's all been framed around cutting costs and driving efficiency is exactly the wrong [approach]."

I think unions can be models for a different relationship with technology—one that emphasises quality, impact, and human-centred work rather than cost-cutting. With this in mind, I repeated Nobel Prize winner Daron Acemoglu’s message: unions need to become technology leaders, not followers. This means engaging critically with AI, understanding its capabilities and limitations, and shaping its development and deployment to benefit working people.

âťť

"We have a role in the union world... to influence the way in which technology is adopted. And we need to be really clear that this technology is being adopted really fast, much faster than we're used to. It means we need to move at a faster rate."

Building international collaboration on responsible AI

The interview concluded with a discussion about cross-border collaboration. Despite differing regulatory environments and political contexts, unions worldwide face similar challenges and opportunities with AI.

The Centre for Responsible Union AI exists precisely to try and increase knowledge-sharing across borders. That’s why connecting to Larry’s invitation to speak was so great. UnionBase are primarily focused on the US, and they’re doing brilliant work in delivering weekly, incisive updates on AI’s impact on labour. They have:

I recommend you subscribe to their YouTube channel or explore their video library, and sign up for their newsletter to stay at the forefront of a worker-centred AI future.

AI transparency statement

This blog post was drafted with support from Claude.ai, following my instructions and using the transcript of the video interview, previous training materials and other relevant data as additional data. I then edited the piece extensively.

Reply

or to participate.