Comparing AI approaches in France and the UK
Gilles Babinet, France’s Digital Champion and Samantha Niblett MP, pick up a conversation that started eighteen months ago, when Gilles visited Parliament to deliver the Digital Leaders AI Annual Lecture, they sat down to compare notes during the AI Public Sector Week, a conversation that gave me new insights.
Here is what stayed with me.
The word sovereignty gets used a lot in AI policy circles, often loosely. But the way Gilles framed it made it land differently for me. The concern is not about closing borders to American technology. It is about what happens to a country’s ability to function when its critical digital infrastructure depends entirely on a handful of external platforms, and the geopolitical weather changes.
France’s response has been to treat compute and data centre capacity as strategic national assets, leaning on its relatively low-carbon energy mix to build infrastructure that keeps economic value inside the country. The UK is moving in a similar direction, with growth zones and distributed data centre investment. Different methods, same underlying anxiety.
One of the things that stood out for me most was how both speakers framed public adoption. Not as a training challenge. Not as a communication exercise. But as a genuinely political and social question: how do you take people with you when technology is moving faster than most people’s ability to make sense of it?
France’s AI cafés were the detail that surprised me most. Rather than running skills programmes or producing explainer campaigns, the approach has been to start conversations. Citizens are not lectured at. They are invited in. The philosophy underneath it is worth sitting with: people should be able to shape how AI fits into their society, not just be informed that it has arrived.
That felt like a meaningfully different instinct to the one that tends to dominate UK policy conversations.
Both speakers were clear on this, and it is something I think gets underestimated. The French example of AI supporting teachers to create adapted learning materials for students with specific needs was not just a good use case. It worked because it was built around what teachers and students actually needed, not around what the technology could do.
The principle is simple to state and genuinely hard to follow: redesign the service around the citizen first, then decide where AI helps. Not the other way around.
There was no pretending here that AI will not disrupt work. It will. But the conversation pushed past the usual reassurances in a way I found refreshing. The point was not “do not worry, new jobs will appear.” It was more specific: the AI infrastructure build itself, data centres, renewable energy systems, technical deployment, creates demand for trade and technical skills. Workforce planning cannot be left to the market alone, especially for people who are least protected when transitions happen.
The word that came up, and that I kept thinking about afterwards, was solidarity. Not productivity. Solidarity. That is a different kind of ambition.
This was the moment in the conversation that I found the most important. The observation that representation in AI development in France is going backwards, not forwards, was not offered as an abstract concern. It was framed as a direct threat to fairness in public services. Who is in the room when these systems are built shapes what those systems do. If that group is narrowing, the consequences are real.
It is easy to treat diversity in AI as a nice-to-have. This conversation made it harder to do that.
What I came away thinking is that the UK and France are not as far apart as their different political contexts might suggest. They share the same underlying questions, about trust, sovereignty, inclusion, and what it actually means to redesign public services rather than just digitise them. Watching two people who clearly respect each other work through those questions in real time was genuinely worth the time.
Watch the full conversation here: https://aipsweek.digileaders.com/talks/comparing-ai-approaches-in-france-and-the-uk/
