How Europe Can Compete With the US and China on AI

Europe can compete in artificial intelligence by focusing on specialist systems for specific industries and tasks, rather than trying to outspend the United States and China on ever-larger frontier models. That is the argument made by Mauro Macchi, identified in the source as Accenture’s EMEA CEO and its digital sovereignty lead.

Macchi says Europe’s industrial history has left companies and institutions with engineering data, designs, maintenance records, research and development, and operational expertise that are difficult to reproduce through computing power or investment alone. He describes “Specialist AI” as systems built around an organization’s or industry’s knowledge, data, workflows, and constraints.

The article cites research showing that one in three organizations exhaust their token budgets before year-end, while less than 20% of that spending is tied to measurable financial returns. An analysis of more than 9,000 tasks found that roughly 90% did not require frontier-model capabilities, according to the article. It argues that companies could reserve the most powerful models for tasks that need them and use smaller, specialized models for other work.

As an example, the article says the European Patent Office processes 400,000 pages a day with a purpose-built model containing one billion parameters, with an error rate below 1%. It argues that Europe’s industrial advantage lies in turning specialized knowledge into AI systems, including local or in-house models for sensitive or regulated work.

The views are Mauro Macchi’s and are not necessarily those of TechRadar Pro or Future plc.