[Japan] Finalisation of "Interim Report" that will serve as basic policy of new AI law
What
On 4 Feb 2025, Japan media reported that the AI Strategy Council had finalised the “Interim Report" of the AI Strategy Council and AI Regulatory Study Group. The Report will serve as the basic policy of a new law aimed at balancing the promotion of AI development with ensuring safety. The new law will apply to both AI developers and businesses that use AI, and is expected to be approved by the Cabinet in late February.
The background to this is that on 26 Dec 2024, the AI Strategy Council had issued for public consultation a draft “Interim Report”. Now, it appears that the Interim Report has been finalised and was published together with the Results of the Call for Opinions on the draft Interim Report.
According to the media, the law will address risks posed by generative AI such as human rights violations and misuse in cyberattacks, the creation of deepfakes used to spread false or misleading information, and crimes such as fraud that could be facilitated by AI.
If there have been significant human rights violations caused by AI, or if there is no improvement [by the companies?] after guidance is provided, the names of thecompanies will be publicly disclosed. But interestingly the law will not contain any penalties due to concerns that excessive regulation may hinder innovation.
The relevant ministries and agencies will be responsible for investigating whether a case is malicious in nature. The criteria for determining if a case is malicious will be outlined in a basic plan to be drawn up after the new law comes into effect.
Below I extract the key / interesting points from the results of the public consultation as well as the finalised Interim Report.
What is the AI Strategy Council?
The Japan government had on 9 May 2023 announced the establishment of an AI Strategy Council comprising experts, researchers, and corporate leaders, to lead the creation of domestic AI policies.
There are few to zero substantive reports in English on what the AI Strategy Council has been working on, probably because their materials - agenda, handouts, minutes - are in Japanese and it is very tedious to get a workable translation.
Results of the Call for Opinions on the “Interim Summary (Draft)”
Key / interesting points from the summary of results:
There was a total of 4,557 submissions. 4,217 came from 1,868 individuals, and and 340 came from 38 organisations.
Many comments from individuals focused on copyrighted works being used without permission, harm from deepfakes, false/misleading information; and called for regulation with penalties.