Europe must build own AI or risk getting cut off by US or China, says ECB’s Lagarde
Christine Lagarde, president of the European Central Bank, has warned that Europe must urgently develop its own artificial intelligence models and build more datacentres to avoid becoming dangerously dependent on the US or China. Speaking in Vienna, she argued that if this dependence is left unchecked, trade partners could gain unprecedented leverage over Europe, potentially threatening to cut off or alter access to AI systems that will soon underpin essential services such as border checks, tax audits, rail networks, hospital care and banking payments.
Lagarde cited stark figures: the US produced 59 notable AI models last year and China 35, compared with just one each from France and the UK, while the US hosts 75% of global AI computing capacity against Europe's mere 5%. She said Europe faces "an awkward choice" between holding back on AI adoption to protect its data and economic autonomy, or adopting it rapidly and risking dependency, and pointed to strained trust with the Trump administration—over tariffs, Greenland and US troop withdrawals—as reason for urgency. She noted AI could boost productivity by up to 4% over a decade if adopted quickly, but warned that Europe's datacentre capacity gap is projected to grow more than sixfold within ten years, while European pension funds' heavy exposure to US tech stocks leaves savings vulnerable to any market correction.
In short
- Lagarde urges Europe to build its own AI to avoid US/China dependency
- US has 75% of global AI computing capacity; Europe only 5%
- Dependency risks giving trade partners leverage in tariff or tax disputes
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Europe should build its own AI models and data centres because systems used in essential services could give foreign suppliers significant influence over access, costs and data. Developing domestic capacity could strengthen resilience and bargaining power, while enabling Europe to adopt AI and capture productivity gains without relying so heavily on US or Chinese infrastructure.
The case against
A rapid push for European AI capacity could require large public and private investment, consume substantial energy and duplicate services that established global providers may offer more efficiently. Europe could protect its interests through diversified suppliers, strong data safeguards and targeted investment, while avoiding a costly race for self-sufficiency that might slow adoption and divert resources from other priorities.
Explained simply
The European Union sets shared rules and policy for its member countries, while each country also makes some decisions about technology and public services. Artificial intelligence systems are built using computer models and large amounts of data, and running them at scale requires powerful computers housed in datacentres.
The United States and China currently have far more AI models and computing capacity than Europe. European governments, businesses and public services may therefore rely on systems and infrastructure controlled by companies abroad.
This matters because AI could help automate work in areas such as healthcare, transport and banking, and may affect productivity across the economy. Reliance on overseas providers also raises questions about access, data control and the resilience of essential services if relations or markets change.
AI Americas Art Asia Business Culture Europe Government Politics Technology World