The phrase Japan’s digital minister chose was deliberately stark. Hisashi Matsumoto warned that the country risks becoming an “AI colony” if it fails to keep pace with the technology, using the term to defend a government-backed bill that would amend Japan’s personal-data protection law to let AI developers use medical and criminal records without obtaining individual consent.

The warning is grounded in a competitiveness gap the government has acknowledged for months. Japan lags not only other advanced economies but also some smaller ones in AI development, by its own assessment, and the gap has been widening year on year, even as the broader race tightens elsewhere and China narrows the US lead to a few percentage points.

Matsumoto’s “AI colony” framing casts that gap as a sovereignty question: a country that cannot build its own AI capabilities ends up dependent on the systems and rules others set.

The bill at the centre of the argument is where the trade-off becomes concrete, and contentious. Easing consent requirements for sensitive categories, medical histories and criminal records, would give Japanese AI developers access to the kind of large, high-quality datasets that train competitive models.

It would also weaken individual control over some of the most sensitive personal information a state holds, which is precisely why proposals of this kind draw scrutiny wherever they appear.

Matsumoto’s case is that the cost of caution is itself a danger. The minister emphasised the urgency, arguing Japan cannot afford to lag, framing the data-access change as a necessary input to closing the gap rather than an erosion of privacy for its own sake.

The counter-position, familiar from data-protection debates elsewhere, is that consent rules exist for the sensitive categories most precisely because the risk of misuse is highest there, the same balance Europe has tried to strike through the EU AI Act.

The bill is one piece of a broader government push. Tokyo is also preparing a large-scale pilot of Gennai, a generative AI platform built for internal government use, planned to reach about 180,000 civil servants across 39 agencies, part of an effort to accelerate adoption inside the state and prod the private sector to invest. The data bill supplies the raw material; the Gennai rollout supplies the demonstration.

The colonial metaphor carries a particular charge in this context. By invoking it, Matsumoto is reframing what could be read as a deregulation of privacy as, instead, a matter of national autonomy, the difference between a country that builds AI on its own data and one that rents capability from systems trained and governed abroad.

Whether that reframing persuades a public asked to give up consent protections over medical and criminal records is the political test the bill now faces, in a climate where surveys already show a wide gap between AI insiders’ optimism and everyone else’s anxiety.

Whether the “AI colony” framing wins the argument is a question for Japan’s legislative process, not a minister’s warning. The bill sets a genuine tension, between the data access developers say they need and the consent protections citizens currently have, that other governments are navigating in their own ways.

Matsumoto has chosen to resolve it in favour of speed, and to name the alternative in the bluntest possible terms. The Diet will decide whether the country agrees.