The Organisation for Economic Co-operation and Development  launched version 2.0 of the Hiroshima AI Process voluntary reporting framework at a Tech7 event held on the margins of the G7 Digital and Tech Ministerial Meeting under the G7 French Presidency in Paris.

As the only global framework for companies to report on their efforts to promote safe, secure, and trustworthy AI, this new and streamlined version is designed to facilitate participation from actors across the AI ecosystem, with a particular focus on small and medium enterprises. It contributes to monitoring the application of the Hiroshima Process International Code of Conduct for Organisations Developing Advanced AI Systems, a central component of the Hiroshima AI Process launched during Japan’s 2023 G7 Presidency.

By contributing to the reporting framework, participating organisations provide comparable information on their AI practices, fostering trust and accountability in the development and deployment of advanced AI systems and helping disseminate good practices. More than 50 companies have already pledged to complete the reporting framework.

Building on the successful first year of the reporting framework, the revisions launched today extend its reach beyond large model developers to the broader ecosystem of AI developers and deployers.

“Enabling a broad range of organisations to share their practices and demonstrate their focus on security, accountability, and transparency will contribute to the adoption of AI across industries, sectors and firms of all sizes,” OECD Deputy Secretary-General Yasushi Masaki said. “This streamlined process will in particular support small and medium enterprises with its improved interface and streamlined approach.”

By aligning the reporting framework with multiple risk management systems, including the Hiroshima Code of Conduct, the OECD aims to promote interoperability and consistency of AI systems around the world.