On February 4, 2025, the Japanese Authorities introduced its intention to place Japan as “essentially the most AI-friendly nation on the planet”, with a lighter regulatory strategy than that of the EU and another nations. This assertion follows: (i) the Japanese authorities’s latest submission of an AI invoice to Japan’s Parliament, and (ii) the Japanese Private Knowledge Safety Fee’s (“PPC”) proposals to amend the Japanese Act on the Safety of Private Data (“APPI”) to facilitate the usage of private knowledge for the event of AI.
The AI Invoice
On February 28, 2025, the Japanese Authorities submitted its “Invoice on the Promotion of Analysis, Growth and Utilization of Synthetic Intelligence-Associated Applied sciences” (“AI Invoice”) to Parliament. If enacted, it could change into Japan’s first complete regulation on AI.
The AI Invoice solely imposes one obligation on non-public sector entities using AI-related expertise: they have to “cooperate” with government-led initiatives on AI. This obligation is predicted to use to entities that develop, present, or use AI-based services or products in reference to their enterprise actions. It’s at present unclear what this obligation to “cooperate” entails. Against this, the AI Invoice requires the federal government to: (i) develop AI pointers consistent with worldwide requirements and (ii) gather info and conduct analysis on AI-related applied sciences. Due to this fact, it’s seemingly that private-sector entities will likely be anticipated to adjust to these government-issued pointers and cooperate with government-led knowledge assortment and analysis initiatives.
Relating to its geographic scope, the AI Invoice doesn’t explicitly present for its software to corporations exterior of Japan. Nevertheless, on condition that the Japanese authorities has emphasised the significance of making use of the AI regulation to overseas corporations in its “Interim Abstract” on AI regulation, it’s attainable that this obligation could possibly be prolonged to corporations working exterior of Japan.
The AI Invoice doesn’t present penalties for non-compliance, furthermore. Nevertheless, the AI Invoice requires the Japanese Authorities to investigate circumstances during which the improper or inappropriate use of AI-related applied sciences has resulted in hurt to the rights or pursuits of people, and to take needed measures primarily based on its findings. Due to this fact, in notably egregious circumstances the place such violations of particular person rights and pursuits happen, it’s seemingly that the Authorities will take measures, similar to issuing steerage for adoption by the related entity or publicly disclosing its identify.
Proposed Modification to the Knowledge Safety Regulation to Promote AI Growth
On February 5, 2025, the PPC proposed introducing exemptions to the requirement beneath the APPI to acquire an information topic’s consent (i) when amassing delicate private knowledge (e.g., medical historical past or felony document), and (ii) when transferring private knowledge to 3rd events.
The PPC is of the opinion that if private knowledge is used to generate statistical info or to develop AI fashions and the outcomes can’t be traced again to particular people, the danger of infringing the rights and pursuits of people is low. Due to this fact, the PPC urged that AI builders ought to have the ability to gather publicly out there delicate private knowledge with out acquiring the consent of knowledge topics if the developer ensures that the non-public knowledge will solely be used for the event of AI fashions. The PPC additionally famous that knowledge controllers ought to have the ability to make such knowledge out there to AI builders with out requiring the consent of the info topic.
Whereas a selected draft modification to the APPI has but to be introduced, the PPC’s proposals are a notable transfer towards the promotion of AI improvement by the comfort of knowledge safety restrictions.
* * *
The Covington group continues to watch regulatory developments on AI, and we usually advise the world’s prime expertise corporations on their most difficult regulatory and compliance points within the EU and different main markets. You probably have questions on AI regulation, or different tech regulatory issues, we’re completely happy to help with any queries.
