Previous G20 presidencies worked to advance AI governance and committed to ensuring its opportunities would benefit all. Given the G20’s track record of advancing AI governance, the Shadow C20 notes with dismay the backsliding shown by this year’s G20 presidency. For the first time since 2013, civil society representatives, including public interest advocacy organizations, academics and independent researchers, have been entirely excluded from G20 participation, while industry leaders and business interests have been granted unprecedented access and influence. Human rights went unmentioned in the 2 September 2026 G20 Innovation Ministerial Statement, and its proposed “light touch” approach rejected enforceable obligations on the most powerful actors in the AI value chain.
The regression toward endorsing industry-led governance reflected in the 2026 G20 Innovation Ministerial statement weakens the G20’s own leadership. As the leaders of the most powerful global economies prepare for the G20 summit in the United States this December, this body remains in a unique position to set AI governance on a more equitable, sustainable, safer course that genuinely centers human rights in fulfillment of its stated AI principles.
In defense of the G20’s own foundational AI Principles, the Shadow C20 urges G20leaders to commit to the following five AI policy pillars:
Pillar 1: Red Line Prohibitions for Systems that Violate Human Rights or Evade Human Control
Pillar 2: Meaningful Multi-Stakeholder Governance with Global Inclusion and Participation
Pillar 3: Equitable Distribution of AI Value across Communities and Geographies
Pillar 4: Privacy, Security and Rights-Respecting Digital Public Infrastructure
Pillar 5: Macroeconomic Coordination on AI, Employment and Economic Disruption