Debate on AI regulations starts with expert panel; Government policy based on ‘maximization of benefits, reduction of risk’

The government began discussing artificial intelligence regulations at an expert panel on Wednesday, sparking a debate on the pros and cons of introducing such rules that will last several years.

The government’s basic policies on AI regulation were presented at the daylong meeting of the AI ​​Strategy Council, including a proposal that major AI developers be legally required to disclose certain information.

The basic policy suggests “maximizing the benefits and reducing the risks” of AI to ensure global competitiveness and improve the quality of life in Japan. Specific risks to consider include human rights violations, security-related crimes and infringement of intellectual property rights.

The government issued non-binding guidelines in April requiring AI providers to take security into account. The basic policy states that appropriate regulation should be considered against “high-risk AI systems and AI that could lead to human rights violations and crimes.” They also note that major AI developers may be categorized as high risk because their products have a significant impact on the public, and that “a legal system to supplement the guidelines may be necessary for public safety and security.”

As for AI-based disinformation, the policy notes the usefulness of “dissemination of technologies that provide information about where the content comes from and its history,” a reference to technologies such as the Originator Profile, which identifies who has a piece of information online created.