Behind the numbers: what are union staff feeling about AI?

Since January 2025, we’ve been inviting union staff to take a survey telling us about their experiences with AI. Over 230 trade union staff from around the world have done so. Two key questions it asks are around the hopes and fears of staff – and the results tell us something hopeful about how staff are approaching the prospect of using AI in their work and personal lives.

The hopes: productivity as a byword for many things

Productivity tops the list of hopes for AI, with 70% of union staff who responded to the survey identifying it as a key aspiration. But what does "productivity" actually mean to union staff? To find out, we need to ask the humans behind the numbers.

Luckily, I get to do that. And, when I talk to union staff in the Responsible AI fundamentals training courses I run, it becomes clear that they aren't simply looking to produce more work. Instead, they're looking for some relief from the burdensome work. Work that they feel takes time away from activities they find more meaningful. As one respondent explained:

So, when union staff hope for productivity gains, they're often really thinking about things like:

• Reducing repetitive tasks

• Creating space for more meaningful, high-impact work

• Maintaining quality while managing increasing workloads

• Finding ways to meet growing demands with limited resources

But alongside these practical hopes lie some more ambitious aspirations. One staff member envisioned AI could "bring about a golden age... Where the 'friction' that we experience working with our data, our email, our technology is smoothed away by intelligent agents." High hopes, indeed!

Beyond productivity: innovation and professional growth

While doing more for members matters tremendously, union staff also express hopes for themselves. Innovation ranks high (43% of respondents), with many staff wanting to be part of more forward-thinking, technologically-adept unions.

For some, they hope AI can be a creative partner for problem-solving: "I am hoping that AI can prompt me to think about other aspects of problems that I hadn't considered," wrote one staff member, while another hoped AI could "generate ideas" and provide "inspiration" to them in their work.

Several respondents mentioned communications as areas where AI might enhance their work. One hoped to "improve our digital offering as a smaller union with a smaller budget," while another specifically wanted help "creating content that can encourage people to join unions and become more active."

The fears, they are many…

There’s a lot of media coverage around AI replacing jobs, and obviously this is a HUGE focus for the labour movement. But when you speak to union staff themselves the fear of job losses ranks relatively low (only 19% cited them).

Instead, their primary fears reflect deeper their values:

AI perpetuating biases or creating disparities (53%): This top concern speaks to unions' fundamental commitment to fairness and equity. One respondent worried that AI would replicate existing societal problems: "My primary concerns about AI are its ability to replicate or entrench existing human problems in an already unequal world, our reliance on its colossal aspirations, and potential unregulated use by anti-liberal autocrats." It isn’t surprising that a union staff are suspicious and see AI through the prism of politics. Given the connections of big tech to some of the most regressive political actors of our times, it is totally right to do so!

Privacy and data security (48%): The second-highest concern reflects fear around the trust relationship between unions and their members. Several respondents specifically worried about sensitive member data being compromised. One staff member expressed concern about "I worry a lot about what kind of member data ends up in these systems where people just bung a load of case files on and ask it to summarise something." Again, a well-placed fear, given that in some AI assistants (notably the free version of ChatGPT) the default is for AI models to be trained on the conversations you have.

Loss of human connection (45%): This concern strikes at the heart of union identity, and has the potential to rekindle debates around organising versus servicing models. Several respondents emphasised the unique value of human connection in union work. One participant feared AI could "produce a featureless, limited, inflexible, 'service' for members." Another worried about "becoming too dependent on AI for decision-making" while one staff member expressed concern that AI might lead to "us becoming less creative." A particularly thoughtful response encapsulated this general concern that AI makes us somehow less human: "I have no real fears, except for the general feeling that we are losing the ability of being critical and creative with our own brains, relying way too much on AI providing us all the answers."

What this shows me: union staff are being thoughtful in their engagement with AI

The survey results – and conversations I have with union staff in the training sessions I run – demonstrate to me that many people in our movement are in a place of thoughtful engagement with AI, a place that is neither enthusiastic adoption or resistant scepticism.

Unlike some sectors that might be seen as rushing toward adoption with a focus purely on efficiency or cost-cutting, union staff are thinking deeply about responsible implementation. In other words, whether and how they can use AI in a way that is aligned with broader social goals.

On the one hand, they:

1. Consistently raise questions about fairness, transparency, and accountability in AI systems.

2. Evaluate potential AI uses based on how they support member needs and union objectives.

3. Express concerns about AI's carbon footprint, reflecting unions' broader commitment to social justice issues including climate action.

Yet on the other hand they:

1. Recognise the need to understand AI to properly advocate for members in workplaces increasingly shaped by these technologies

2. See potential for AI to address longstanding challenges in the union movement, that have limited our impact

3. Express genuine enthusiasm for AI's potential to enhance rather than replace human-centred union work

4. Want to experiment responsibly rather than simply adopt blanket prohibitions or restrictions

This balanced approach suggests to me that union staff can offer a powerful and distinct perspective on AI adoption. One that values both innovation and ethical consideration. One that seeks a middle ground between uncritical adoption and wholesale rejection.

And it reinforces my belief that unions have an important role to play not just in responding to AI but in helping shape its development and implementation. By bringing their unique perspective on fairness, equity and collective wellbeing, unions can help ensure that AI serves rather than undermines worker interests.

This blog is based on a survey undertake in 2025 of 232 union staff conducted by The Centre for Responsible Union AI at Unions21. The blog was written with the support of Claude.ai for drafting and identifying key quotes from contributions to our Responsible AI fundamentals training courses.

Image credit: Yutong Liu / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/