The Hiroshima AI Process (HAIP) is a global transparency voluntary framework for advanced AI -a voluntary mechanism designed for organisations that want to lead the conversation on trustworthy AI rather than follow it.
Managed by the OECD (commissioned by the G7), it enables organisations to demonstrate leadership in responsible AI by making governance practices visible, comparable, and credible across borders.
From principles to practice—and proof
Many organisations already have AI principles and internal policies. HAIP asks a more demanding question: How are those principles implemented in practice?
Through a structured but flexible reporting framework, organisations publicly describe how they identify and manage AI risks, allocate responsibility, handle incidents, and consider broader societal impacts. The focus is practical and operational—not aspirational.
Crucially, HAIP reports are publicly accessible. Organisations can browse submissions from peers across sectors and jurisdictions, creating a valuable benchmark for shaping internal policies, governance models, and risk-management practices.
With HAIP, transparency becomes not just a signal of trust, but a source of collective learning.
A strategic complement to regulation
For European organisations, HAIP sits naturally alongside regulatory frameworks such as the EU AI Act. While the AI Act sets legal requirements, HAIP is a voluntary reporting framework that enables organisations to demonstrate governance maturity, foresight, and accountability beyond minimum compliance.
Participating in HAIP allows organisations to show that they are not only meeting obligations but actively contributing to the development of global norms for responsible AI.
Visibility without posturing
HAIP’s tone is deliberately non-punitive. Organisations are not ranked, scored, or “named and shamed”. Instead, reports are published in an OECD-hosted repository that builds a shared, evolving picture of how AI governance is implemented in practice.
Organisations that participate may also use the HAIP brand to signal a serious commitment to transparency and responsible AI. This is not a claim of perfection, but a willingness to engage openly with complexity, risk, and accountability.
Why this matters for the AI Alliance community
Responsible AI leadership is no longer defined by principles alone. It is defined by transparency, accountability, and the confidence to show how governance works in practice. HAIP provides a credible global platform to do exactly that, and it is therefore a recommended tool for the AI Alliance community.
- Prisijunkite, kad galėtumėte skelbti komentarus.