A PYMNTS Company

Japan Considers Regulation of AI Developers

 |  May 7, 2024

Japan is contemplating the regulation of large domestic and international artificial intelligence (AI) developers. According to Nikkei ,this move aims to curtail potential risks to society, including the dissemination of misinformation and involvement in criminal activities. The decision follows in the footsteps of the United States and the European Union, which have already implemented similar measures.

Historically, Japan has relied on a system of self-regulation by companies, guided by government-issued AI guidelines, to foster growth in the sector. However, concerns over the misuse of AI technologies have prompted the government’s AI strategy council to initiate discussions on the development of a legal framework. These discussions, slated to commence in May, will explore the feasibility and implications of imposing regulations on AI development.

While acknowledging the potential benefits of AI, authorities remain wary of its misuse. Of particular concern is the use of generative AI in spreading misinformation and facilitating criminal activities.

The upcoming deliberations will evaluate the pros and cons of implementing laws and regulations governing AI development, drawing insights from regulatory frameworks established in the United States and Europe. Central to these discussions is a draft proposal presented by Japan’s ruling Liberal Democratic Party (LDP) in February.

The LDP’s proposal primarily targets large-scale AI developers, including entities like OpenAI, the developer of ChatGPT. Key provisions of the plan include mandatory third-party verification of safety protocols for high-risk AI development and the compulsory sharing of risk-related information with the government. Additionally, developers would be required to submit regular compliance reports, with penalties imposed for violations.

However, the LDP has yet to reach a consensus on the definition of “large scale” developers or whether such designations should be unilaterally determined. The proposal also leaves certain details to be hashed out by the private sector, underscoring the delicate balance between regulation and fostering competitiveness.

Read more: Meta & OpenAI CEOs Back EU AI Regulations

The envisaged regulations are slated to be incorporated into the government’s overarching economic and reform policies, set to be formulated in June. Subsequently, a bill outlining the regulatory framework is expected to be introduced during the regular session of the Diet in 2025.

Nevertheless, some within the government remain cautious about deviating from the current policy of self-regulation by companies. Companies like Google have voiced concerns that excessive regulatory measures could stifle innovation and hinder the adoption of generative AI technologies.

Internationally, the European Union has emerged as a frontrunner in AI regulation, having passed comprehensive legislation in March mandating monitoring and transparency measures based on risk assessments, with penalties for non-compliance. In the United States, similar initiatives were introduced in 2023 through executive orders, mandating safety evaluations for select AI developers. Meanwhile, China has implemented stringent measures, including bans on the development of generative AI perceived as threatening national security.

Source: Asia Nikkei