Alongside the proliferation of artificial intelligence (AI) technologies, particularly machine learning, generative AI capable of producing diverse content has gained widespread use, raised expectations for improvements and innovations in various aspects of life and work. However, it has also underscored the need for appropriate design, development, deployment, and utilization of AI. Given that AI evolves and operates within society, there is a pressing need to establish comprehensive AI governance that encompasses AI developers, providers, users, public institutions, and the society at large. Emphasizing transparency and equity, the call for a shared framework that fosters innovation while mitigating risk is important.

To advance the responsible deployment of AI, it is imperative to delineate the responsible actors and formulate appropriate measures.

Particularly, in the continuum of processes spanning AI design, development, provisioning, and utilization, where various organizations and individuals may be involved, the locus of responsibility can become ambiguous. Consequently, in interorganizational transactions encompassing AI development to provisioning, ensuring appropriateness through contractual agreements is important. Furthermore, monitoring mechanisms should be established to ensure proper transactions. However, in transactions between AI providers and consumers, providers should not only take suitable preventive and corrective measures, but AI users can also leverage governance through disciplines other than regulations, such as market dynamics, investments, and reputation, by acquiring appropriate literacy. Additionally, it is advisable to consider establishing remedial measures such as compensation systems when accountability is unclear. Considering that the AI lifecycle extends beyond national, regional, and organizational boundaries, it is essential to promote discussions that enhance transparency regarding the responsibilities and measures of these stakeholders. The details of this presentation can be read in the policy recommendation published in September 2023 (https://ifi.u-tokyo.ac.jp/en/news/119....

Learn more about the AI Forum 2023 hosted by Microsoft Research Asia in collaboration with The University of Tokyo: https://www.microsoft.com/en-us/resea...