A milestone in international AI transparency: The OECD publishes initial submissions from the G7 Hiroshima AI Process Reporting Framework
Today marks an important milestone in international AI governance as the OECD publishes the first round of reports submitted through the G7 Hiroshima AI Process Reporting Framework. A diverse group of 19 organisations from around the world has provided detailed information on their practices for developing advanced AI systems in a safe, secure, and trustworthy manner.
A significant step for global AI transparency
The publication of these reports reflects the first tangible outcomes of the reporting mechanism launched in February 2025. This initiative, developed at the request of G7 leaders, offers unprecedented transparency into how leading AI developers are implementing risk management practices aligned with the Hiroshima AI Process International Code of Conduct for Organizations Developing Advanced AI Systems.
Since the launch, organisations from diverse geographies and of various sizes have submitted reports, demonstrating the global reach of this initiative. Participants include:
- Major AI developers
- Established technology companies
- Emerging AI developers
- Independent experts and academic institutions
What these reports tell us
The published reports provide valuable insights into how organisations:
- Conduct AI risk assessments
- Manage potential safety and security vulnerabilities
- Implement governance frameworks for advanced AI systems
- Share information about incidents and best practices
This information is now publicly available at transparency.oecd.ai, providing an unprecedented resource for understanding the state of AI governance practices globally.
Join the movement: help shape a safer AI future
This publication marks the beginning of an ongoing effort to promote transparency and accountability in AI development. The reporting framework will continuously accept submissions on a rolling basis, with participating organisations encouraged to update their reports annually.
Organisations developing advanced AI systems that have not yet participated are encouraged to submit reports through the online platform. The OECD will analyse the submitted reports and publish insights about emerging best practices on the OECD AI Policy Observatory in June.
The publication of these reports signifies a substantial advancement in the implementation of the G7 Hiroshima AI Process and exemplifies how international cooperation can foster progress in AI governance.
👉 Access all published reports and submit your organisation’s report: transparency.oecd.ai
List of organisations that have submitted a report:
