The European Union is putting $11.4 billion into seven AI "gigafactories," according to a report from Memeburn surfaced via Google News.
The term borrows deliberately from car manufacturing. A gigafactory is a very large industrial plant, and applying the word to artificial intelligence signals how the technology is being treated: not as software written in an office, but as something that requires land, power, cooling and enormous quantities of specialized chips. Memeburn frames the investment as evidence that AI is becoming heavy industry.
That framing is the heart of the story. For most of the past decade, the strategic question in AI was who had the best researchers. Increasingly, the question is who has the physical capacity to run the models — the compute. Building that capacity is slow, expensive, and geographically fixed, which makes it a matter for governments rather than individual companies.
The EU's move follows a broader pattern of states treating advanced computing as infrastructure on par with roads or electricity grids, and it reflects a concern that Europe has been dependent on capacity built and controlled elsewhere. Public money at this scale is an attempt to change that.
A caveat on detail: the available source material is a single headline-level report. It does not specify where the seven sites will be located, over what period the money will be spent, which companies or chip suppliers will be involved, or how much of the total is new funding versus previously announced commitments. Those specifics will determine how much the announcement actually changes on the ground, and they are worth waiting for before drawing firm conclusions.
Why it matters: whoever controls the factories and data centers that make AI physically possible will shape who gets to use the technology, on what terms, and at what price — and Europe has just made an expensive bid not to be left renting that capacity from someone else.