Skip to content
Unions 21
| Blog post

Getting ready for government transition. Can AI help unions?

By Jonathan Tanner, Rootcause | 4 min

Do you ever feel like it’s impossible to keep up these days? I hate to break it to you but information overload is real and it is going to get worse. If you’re anything like me then you’ll already have gotten used to leaving dozens of well-intentioned newsletter subscriptions and podcast episodes languishing unopened in perpetuity. 

Now with the rise of AI-generated ‘slop’ across the internet, all signs point towards a future where there’s more and more stuff to read, podcasts to listen to and videos to watch in order to feel you’re staying on top of understanding the world. The amount of this content which is created or packaged by AI will grow and the expertise or information savvy that you need to separate signal from noise is likely to become more sophisticated.

This sense of drowning in information is challenging for all of us professionally. We want to stay up to speed but it’s almost impossible to find the time. Thankfully, there are solutions to this problem beginning to emerge, but as is often the case they are not yet perfect - especially because many of them involve the use of AI. 

The idea of using AI to solve problems that AI will help create can feel frustrating, especially when many of the companies creating and deploying AI technology are not acting in a way which is ethical or transparent. There is also justified unease around the labour market impact of these technologies. 

Yet for many of those in the union movement there is potential for AI to help address some significant day to day challenges. If the next month sees a new Labour government elected there is going to be a flurry of consultation and legislation which will require engagement from across the board. This will be a big lift for many people already feeling busy and overwhelmed. 

At Rootcause we work with progressive organisations to help them succeed in digital information environments. We can see several ways that unions might benefit from exploring the potential of AI to help them prepare for and respond to a new government. 

One such idea is the use of AI to generate summaries of new legislation and consultations. This is something we did for the European Union’s Artificial Intelligence Act where we tracked new amendments placed by politicians from across the political spectrum to analyse the intent and possible consequences of proposed legislative change. This enabled us to identify which politicians were working together around which specific issues (and to see which corporate actors were aligned with those groups). 

To do this, we scraped the European Union's legislative Observatory platform which provides all of the text to amendments as they are laid. This was some of the more technical work we had to do.

We asked GPT to identify the amendment to the original text in order to identify the implications of the amendment.  It made a snap judgement on what the amendment was trying to achieve and categorised it. We then asked for a summary of the intent and whether it was a substantive or technical change to the legislation that was being proposed.  

Finally, we asked GPT to score the impact of the amendment on a series of scales including openness to new technology, attitude to technology companies and support for human rights. 

We felt that what we did here would probably save a lot of time that's often spent reading through and thinking through implications of big, dense documents 

The summarising potential of AI is matched by its ability to augment strategic thinking - analysing draft consultation responses for completeness, unintended consequences or logic gaps and adopting different personas to critique emerging arguments are all in its wheelhouse. 

So too is the ability to identify new insight from union members, using chatbots to engage in fact-finding conversations that approach things different to traditional survey or focus group work and using the analytical power of AI to rapidly identify topics and trends in the responses.  

These are exciting ideas that offer the potential to bolster efforts to influence a new government at a time when capacity is already stretched and the ability to keep pace with the machinery of government and resources of corporate interests will be tested. 

A striking feature of the last decade or so has been the willingness of regressive actors to experiment with new technology in order to pursue their agenda. These actors have had some notable successes. It would be a good time for the trade union movement to score some victories of its own. Getting to grips with AI - for better and for worse - offers a way to do just that. 

More ideas