Japan's digital chief pushes consent free AI data access to avoid dependence

Japan’s Digital Minister, Hisashi Matsumoto, said on Friday that the country risks becoming what he calls an “AI colony” if it can’t close the gap with global competitors.

Matsumoto used the phrase “AI colony” while defending a bill that would allow AI developers to train models on sensitive personal data without consent. The data includes medical records and criminal histories.

“I hope many Japanese people understand that we need to press ahead with AI development, or we’ll end up becoming an ‘AI colony,'” Matsumoto said at a press briefing, according to Jiji Press.

Japan’s AI data bill is splitting the parliament

The amendment passed the lower house last week. Now it’s in the upper chamber, where opposition lawmakers are standing against it.

Their objection is that the changes open the door to data breaches and gut privacy protections.

Matsumoto said the revision won’t lead to the leaking of personal information. The bill limits expanded data access to statistical use cases related to AI development.

Japan’s AI spending gap is massive compared to the US and China

From 2019 to 2023, the US spent ~$329 billion on local AI research, while China spent ~$133 billion. Japan has spent only about $10 billion on AI, which is a big difference that led to a new policy.

To speed up AI development, Tokyo changed subsidy rules, provided funding, and pushed for legal changes. It has also tried to get American tech companies to do business in Japan.

Microsoft and OpenAI have both deepened their collaboration with Japan under the US-Japan security alliance framework.

Moreover, Japanese officials are backing local AI projects. SoftBank, Sakura Internet, and domestic chipmakers got government support to build local AI models and computing infrastructure.

OpenAI visited Japan in late May to pitch GPT-5.5 Cyber, a cybersecurity-focused AI system, to government officials and private companies. Paul Nakasone, an OpenAI board member and former head of US Cyber Command, said the company discussed defense measures across 15 critical sectors with Japanese officials.

But not everyone in Tokyo thinks building a fully domestic AI stack is realistic.

When the Ministry of Economy, Trade and Industry proposed developing a Japanese equivalent of ChatGPT using public funding, some ruling party lawmakers called the plan reckless. They said Japan doesn’t have the resources to compete with American and Chinese rivals.

METI dropped that specific goal. But the government is revising its Basic AI Plan this summer, with draft language expected to strengthen sovereignty provisions tied to national security.

A competing view is gaining traction within the ruling Liberal Democratic Party’s Digital Society Promotion Headquarters. Secretary General Akihisa Shiozaki said in May that Japan should focus on diversifying its AI suppliers rather than building sovereign systems from scratch.

“What matters most is ensuring autonomy without becoming dependent on any single country, company, or provider,” Shiozaki said.

The developed world shares Matsumoto’s concerns. Earlier this week, the European Union announced a technology sovereignty package aimed at strengthening domestic cloud, AI, and semiconductor industries while reducing dependence on American tech companies.

If you're reading this, you’re already ahead. Stay there with our newsletter.