Partnership on AI

Can you tell the difference between a real Thanksgiving moment and an AI-generated one? Learn what signs you should be on the lookout for, then put your skills to the test with our quiz.

“[AI] is a product of the choices we make as individuals, as companies, and as governments about the way in which we want to use it.” - Rebecca Finlay PAI CEO and McGill alumna, Rebecca Finlay, shares how PAI is working with Partners around the world to make AI more transparent, accountable, and aligned with the public good in this feature story from McGill: https://buff.ly/iIrs2KT Tom Kertscher Alyssa Bistonath

2025 may be dubbed the year of agentic exploration, but adoption of AI agents is slowed by reliability and security challenges. This gives policymakers and researchers a critical window to prepare thoughtfully for how these systems will shape society. Partnership on AI’s agenda, Preparing for AI Agent Governance, provides a roadmap for building the evidence policymakers will need. It highlights: 1. Why uncertainty about AI agents makes research, not premature regulation, the top priority 2. The role of monitoring, sandboxes, and testbeds in generating evidence under regulatory supervision 3. 12 top-level and 45 sub-level research questions for policymakers and researchers to explore This work builds on PAI’s recent research on failure detection and global governance, and is designed to support evidence-based policymaking as AI agents develop. 📄 Download the full agenda here:

AI agents go beyond traditional systems — autonomously setting goals, executing complex tasks, and even writing and running code. They raise governance challenges that transcend national borders. Addressing these challenges requires more than domestic regulation, it requires global governance. Partnership on AI’s paper, AI Agents & Global Governance: Analyzing Foundational Legal, Policy, and Accountability Tools, examines how international law, non-binding global norms, and accountability mechanisms can be leveraged to manage cross-border risks, and where these tools must be strengthened. 📄 Read the full analysis

Building an AI future that benefits society takes courage, collaboration, and enduring commitment. PAI’s ChangeMakers Award celebrates leaders who exemplify these qualities and have left a lasting mark on the field of responsible AI. This year's recipients - Jerremy Holland (Apple) and Francesca Rossi (IBM)- have played a critical role in guiding PAI’s evolution and advancing the field at large. From championing equity and justice in AI to fostering cross-sector partnerships, their work exemplifies what it means to lead with purpose. We’re proud to celebrate these outstanding ChangeMakers and the impact they’ve had on our community and beyond. https://lnkd.in/gZ3nZebc

AI is reshaping how companies create value, compete, and manage risk, but there are significant gaps in how they report AI-related impacts, risks, and opportunities. PAI's new landscape analysis examined how 50 of the world's largest companies currently disclose AI’s material impacts in their financial and sustainability reports, identifying where transparency exists and where it falls short. We found that even among the world’s major companies, the quality, quantity, and location of disclosures vary significantly. While best practices have emerged, there are significant gaps in how companies report AI-related impacts, risks, and opportunities. Download the full analysis here: https://buff.ly/Ij8HNxT Thank you to the authors of this timely analysis: Stephanie Bell, Dunstan Allison-Hope, and Mary Beech

The global conversation on AI governance is moving fast, and clarity has never been more important. We’re inviting public feedback on a Zero Draft report that introduces three tools to help policymakers, developers, and advocates navigate this complex landscape: ☰ Governance Stack - Layers AI laws, standards, and principles from broad values to technical details 📍 Governance Map - Charts AI governance activity to highlight strengths and gaps 🪝 Anchors & Hooks - Shows how governance efforts connect and build on one another Published jointly with Microsoft, this resource is open for public comment. Explore the tools and share your feedback.

How can we build AI systems that reflect the needs of the people they impact most? Earlier this year, we hosted a conversation on the value of participatory public engagement in AI development with Tina M. Park and members of our Global Task Force for Inclusive AI. Grounded in insights from our Guidance for Inclusive AI, this discussion explores why and how developers and deployers should engage directly with the public and what’s at stake if they don’t. 📺 Watch the full recording on YouTube

Next week! We’re proud to be part of #DTC2025 with Mila, OECD.AI, World Privacy Forum, The Centre for Media, Technology and Democracy, KI Park, and Bertelsmann Stiftung. Our CEO, Rebecca Finlay, and International Policy and Government Lead, Talita Dias, will share insights from the PAI community on trustworthy and responsible AI. Register now ➡️ https://buff.ly/e2aQVBS