AI policy needs evidence. But the evidence does not come from one panel, one region or one definition of risk. At #DigitalCooperationDay, four international AI evidence-building initiatives came together to compare where their findings on AI opportunities, risks and impacts converge, where they differ, and what is still missing. Yoshua Bengio and Maria Ressa, Co-Chairs of the Independent International Scientific Panel on AI, were joined by Kwan Yee Ng 吴君仪 of the International Scientific Report on the Safety of Advanced AI, Kalika Bali of the Global South AI Safety Report, and Adam Gleave of the EU AI Act Scientific Panel. Gary Marcus offered framing remarks, @Renata Dwan moderated the discussion, and Urvashi Aneja brought perspectives from the Global South Network for Trustworthy AI. The question running through the discussion: how can scientific evidence become more useful for real governance choices across countries with very different capacities and resources? 🔴 Watch: https://lnkd.in/gtZ8qFgx #DigitalCooperation #DigitalCooperationDay #AIGovernance
An important question, particularly the emphasis on evidence coming from multiple regions, methodologies and definitions of risk. Global AI governance cannot simply export the assumptions of whichever jurisdictions have the greatest compute capacity. We also need mechanisms that can translate diverse evidence into operational constraints while preserving context, uncertainty and traceability. That is part of what we are exploring through Inter Dimensional Computation | IDC™: moving governance closer to the computational layer, where decisions can be evaluated across temporal, contextual, ethical and uncertainty dimensions, with their lineage preserved through RealTime Transparency Scroll | RTTS™. Scientific plurality matters. The difficult next step is turning that plurality into systems that can actually reason with it. 🌍 #AIGovernance #ResponsibleAI #AITrust #AITransparency #DigitalCooperation #GlobalAI #AISafety #DigitalGovernance #InterDimensionalComputation #AuroraEthica
A valuable reminder that effective AI policy must be grounded in diverse, credible and continuously evolving evidence. Bringing together different perspectives can help identify both shared challenges and important gaps in our understanding of AI’s impacts.
This raises an important distinction between policy evidence and execution evidence. Policy can define what should be permitted or prohibited, but policy alone does not necessarily produce evidence of what an AI system actually did. Execution does. For increasingly capable Agents, the execution layer can continuously generate evidence about: what was proposed, what was authorized, what was actually executed, under which context and constraints, and what happened as a result. This creates an important feedback loop: Policy → Execution → Evidence → Evaluation → Policy Evolution And this is where the “unknowns” become particularly interesting. Evidence is not only useful for proving that a known rule was violated. It can also reveal behaviors and execution paths that were never anticipated when the original policy was written. In other words: Evidence can reveal the unknowns that policy did not know it needed to govern. For AI governance, this may be an important bridge between policy on paper and safety in operation.
AI governance cannot be built around evidence produced from only one region or one understanding of risk. What matters is whether evidence reflects the realities of countries with very different capacities, resources and digital contexts. The Global South perspective is therefore not an optional addition to AI governance, it is essential to building policies that are relevant beyond the countries developing the technology.
AI Policy needs Evidence. What is Policy? How to build a Culture of Innovation? How to define Scientific Evidence? What are the core characteristics of Scientific Evidence? Core Criteria for Policy Relevance: How to define Trustworthy AI? Real Governance Choices across countries and regions with different capacities and resources: What is AI Architecture? About AI Governance? How to define Sovereign Talent? What is the Global South Perspective? BEUNS LLC at the 2026 UNGA-81 Global Events from September 8-28 in New York City, USA: We put PEOPLE first. BEUNS LLC is not only a global enterprise but also the Human-Centered Paradigm AI HUB, bringing together the best-in-class experts from 193 countries. Global Leadership Spotlight: At BEUNS LLC, we take decisive action, deliver historic results, and achieve monumental milestones. We unlock human potential and unleash productivity.
United Nations Office for Digital and Emerging Technologies GBS-GLOBAL is framework defined by the world system currently in to make it easier for world leadership requirements. Download the detailed work on my homepage as it might assist with the question posed. https://capstoneprojectsjs.netlify.app/
Spiral With Fredah LLC endorses the UN Global Digital Compact.
einstein started chat e mc squared 1905 actually only well engineered kayer 5 ai can save us .more at x yunuslab
I am doing Action now!! Only Action!
AI Policy needs Evidence. What is Policy? How to build a Culture of Innovation? How to define Scientific Evidence? What are the core characteristics of Scientific Evidence? How to define Trustworthy AI? Real Governance Choices across countries and regions with different capacities and resources: What is AI Architecture? About AI Governance? How to define Sovereign Talent? BEUNS LLC at the 2026 UNGA-81 Global Events from September 8-28 in New York City, USA: We put PEOPLE first. BEUNS LLC is not only a global enterprise but also the Human-Centered Paradigm AI HUB, bringing together the best-in-class experts from 193 countries. Global Leadership Spotlight: At BEUNS LLC, we take decisive action, deliver historic results, and achieve monumental milestones. We unlock human potential and unleash productivity.