The Japanese government is considering introducing legally binding regulations on developers of large-scale artificial intelligence systems to ensure they implement measures to address disinformation and other risks, sources close to the matter said Monday.
While the government had previously leaned toward making such measures voluntary, it has recognized the need for penal regulations, similar to moves made by the European Union and other countries, amid concerns over potential AI misuse.
The government plans to convene a council of AI experts to discuss the issue and is considering incorporating the new regulations into economic and fiscal management policy guidelines to be compiled around June.
Japan will soon release guidelines listing 10 principles, including "human-centeredness" and safe AI use.
According to a draft released last month by a project team of the ruling Liberal Democratic Party, businesses developing advanced technologies, such as the generative AI chatbot ChatGPT will be designated as "AI foundation model developers."
Companies involved in the use of AI in high-risk areas will be required to conduct internal or external safety verifications and share risk assessments with the government.
Government-designated developers will also be requested to report their compliance status to the government or third-party agencies. In cases of non-compliance, the government would be able to request reports or conduct on-site inspections, as well as impose fines and other penalties for violations.
The European Parliament earlier this month passed the world's first comprehensive artificial intelligence act, expected to take effect from 2026 and will impose significant fines for violations.
Related coverage:
Amazon to invest 2 tril. yen in Japan's cloud business amid AI boom
Japan plans legal revision for quicker deletion of defamatory posts
G7 sets out guiding principles for AI developers to address risks
