Few topics generate more confident predictions than artificial intelligence. It will transform healthcare, reinvent work, reshape geopolitics. And running through almost all of it is a competitive logic: who has the largest models, the most data, the greatest compute. Europe, in this telling, is a step behind. Francesco Ferrero has a different read.
"The narrative is that Europe is lagging behind the US and China in the bigger-is-better AI competition. But that is not a big deal, because it is the wrong competition."
The wrong race
Scale, in the current AI paradigm, means building ever-larger models on ever-larger datasets, requiring ever-greater compute. The results can be impressive. The costs are harder to brush aside.
"The bigger-is-better approach is not sustainable from either the economic, societal and environmental perspectives."
The economics alone are telling: only a handful of organisations on earth have the infrastructure to train frontier models. For everyone else — the hospitals, the public agencies, the small businesses that might actually benefit most from AI — those systems remain inaccessible or unaccountable, things that happen to them rather than tools they control. And the environmental footprint of large-scale AI is no longer a footnote; it is a genuine constraint on how far this approach can go.
Europe, Francesco Ferrero argues, need not chase that particular horizon. Its advantage lies elsewhere: in building AI that is trustworthy by design, efficient with resources, and genuinely oriented around the people who use it. In a world of growing instability and resource pressure, that is a more durable bet than raw scale.
What responsible AI actually means
"Responsible AI is an artificial intelligence that is factually correct, that can explain to the people the decisions that it takes that affect them, that is not biased and doesn't discriminate, and that is frugal."
Each element of that definition has teeth. A system that cannot explain its decisions to the people affected by them is not accountable in any meaningful sense, it is simply authoritative. A system that encodes bias will reproduce it, reliably, at scale, in the decisions that shape people's access to jobs, credit, healthcare. And a system that requires enormous resources to run is one that most organisations will depend on without understanding, rather than deploy with control.
At LIST, these are design requirements, not aspirations. The research group develops technologies that make frugal, explainable and fair AI buildable in practice.
Tools for a different kind of leadership
The LIST AI Sandbox was built around a question that sounds simple but rarely gets asked: how do you actually know whether an AI system you are using is safe, fair, and doing what you think it is? For most organisations, the honest answer is that they do not.
"The LIST AI Sandbox allows everyone to test the AI systems that they use."
The Sandbox addresses that gap directly, giving any organisation the means to stress-test AI systems against real conditions before they go anywhere near a consequential decision. In healthcare, in justice, in financial services, that kind of independent scrutiny should be routine. It largely is not.
The Luxembourg AI Factory operates at a different scale. Where the Sandbox offers a testing environment, the Factory is designed to reshape how AI gets adopted across the whole economy, with a particular focus on the businesses most likely to be left behind.
"The Luxembourg AI Factory will accelerate the adoption of ethical AI by all the citizens of Luxembourg and Europe, and especially SMEs and startups."
As a founding partner, LIST guides organisations from first ideas through to working deployment, connecting the research, the tools, and the sectoral expertise that turns responsible AI from a principle into something an SME can actually build with. The goal is not to produce another tier of AI inaccessible to most organisations. It is the opposite.
Technology people can embrace
Francesco Ferrero returns, at the end, to what is perhaps the oldest question in the history of technology: what happens when the people it is supposed to serve do not trust it?
"Frugal and human-centred AI is the right approach for Europe because in a world that is increasingly affected by conflicts and scarcity of resources, we need a technology that the people can embrace instead of one they will fight in the long term."
AI systems built without explainability, without fairness, without any mechanism for accountability, generate resistance. Not immediately, perhaps, but steadily — from citizens who experience their effects, from regulators who are asked to oversee them, from societies that find the terms of the technology were set elsewhere, by others, for other purposes.
The harder work, and the more important one, is building AI that earns trust. Can it reinforce democracy instead of eroding it? Can it make organisations more competitive without making people more vulnerable? Can innovation and ethics travel together?
