Annual Report 2025

Partnership on AI (PAI) is a nonprofit global community of academic, civil society, and industry organizations who come together to address the most important questions in AI. In 2025, we advanced the safety and responsibility of AI systems. We broadened our global community, expanded our expertise, prepared for emerging technology, and shaped conversations around AI’s impacts on people, trust, and inclusivity.

ANNUAL REPORT 2025

From Our CEO

In the past year, the use cases and applications of AI have expanded significantly–affecting everything from enterprise workflows and worker protections to social relationships and mental health. To meet this moment, Partnership on AI expanded our reach across sectors, industries, and continents to activate a community-wide effort to prioritize the development of AI in the public interest.

We developed resources to help technologists and policymakers better assess and govern AI agents. We applied our track record in synthetic media responsible practices to launch a new focus on AI’s impacts on human connection and trust. And we established a foundation for the current state of formal reporting to understand the impacts of AI. Today, we are a trusted source for industry and policy guidance and accountability.

Over the past year, we set the stage for greater global impact. We integrated diverse insights from across our network into international standards and best practices at key events such as the France AI Action Summit in Paris, the Global Summit on AI in Africa in Kigali, and the Asia Tech conference in Singapore. With the addition of 13 new partners, two new board members, and the launch of a new European Steering Committee, we are building the world’s first inclusive platform for responsible AI.

As we enter our second decade in 2026, our collective impact is clear and our work is urgent. Thank you to our staff, board, partners and collaborators. Together we are building a future where humanity develops AI to contribute to a more just, equitable, and prosperous world.

In partnership,
Rebecca Finlay
CEO

PAI by the Numbers

141

partners

countries

convenings

1530

attendees from

countries

new resources

areas of work

29k

social media followers

267,133

organic social
impressions

800M+

earned media
impressions

87,397

website visitors

original blogs

1594

downloads of
PAI resources

Media Appearances

“Don’t miss out” replaces “doom is nigh” at Paris’ AI summit

PAI CEO Rebecca Finlay told Axios that important discussions about safety and responsibility were taking place, even if the event’s official actions don’t provide tangible progress.

The Emerging Agentic Enterprise: How Leaders Must Navigate a New Age of AI

Rebecca Finlay argues that companies need to be more transparent about their AI failures.

Unlock Scale with Inclusive AI

Rebecca Finlay tells how AI will only deliver competitive advantage when it actually works for people and society.

The Media Show: Journalists in Gaza, AI Avatar, Housing Journalism, True Crime

PAI’s Claire Leibowicz weighs in on a journalist’s posthumous “interview” with an AI-generated avatar of a school shooting victim.

Meeting the Paris Summit Goal for ‘More Inclusive’ AI Means Getting Stakeholder Engagement Right

PAI’s Tina Park reflects on the Participatory AI Research & Practice Symposium (PAIRS) that preceded the Paris AI Action Summit.

Rebecca Finlay at the PAI Partner Forum in San Francisco

Fostering a Global Community

In 2025 we strengthened connections amongst our global community at events around the world.

Our team has crossed the globe, connecting with partners and collaborators at key events this year, from the AI Action Summit in Paris, where we discussed trust and the need for a global AI assurance ecosystem, to FAccT in Athens, where we exchanged ideas on Participatory Engagement. Our presence at the Global AI Summit on Africa in Kigali, allowed us to take part in meaningful conversations on building a collective action plan for AI across the continent. We brought together our community of Partners, policymakers, and other stakeholders for a series of events alongside the UNGA in New York, to discuss agents and the need for a more robust global AI assurance ecosystem. We closed out the year with our 2025 Partner Forum in San Francisco, fostering conversations with our community centered around trust, the future, international perspectives, and agents.

“Joining the Partnership on AI strengthens our commitment to fostering collaborative, innovative solutions that align AI development with societal benefit, amplifying our impact through PAI’s global community.”

Brian Tse
Founder and CEO
Concordia AI

Stephanie Bell at the IMF Spring Meetings, Washington, DC

Felecia Webb at the Deep Learning Indaba, Kigali, Rwanda

Sonam Jindal at the AI Action Summit, Paris

Stephanie Ifayemi (left) and Rebecca Finlay (far left) at the Global AI Summit on Africa in Kigali, Rwanda

Expanding Expertise and Leadership

In 2025 we expanded our expertise and leadership in AI across the globe, welcoming new board members, partners, and advisors.

PAI Changemakers

Recognizing the most influential leaders in AI, PAI’s Changemaker Awards, recognized the leadership and impact of Jerremy Holland, Director of AI Research at Apple and PAI Board Chair and Francesca Rossi, IBM Fellow, Global Leader for Responsible AI and AI Governance and Founding PAI Board Director.

New Board Members

We welcomed new members Christina Colclough and Helen King to the PAI Board of Directors.

Christina Colclough
The Why Not Lab

Helen King
Google DeepMind

SAIGE Council

We established a global standing advisory body of interdisciplinary researchers and experts to act as advisors, ensuring PAI’s work meets the highest standards of independence and rigor while remaining attentive to external and emerging trends.

European Steering Committee

We launched PAI’s first international steering committee, our first step in building stronger connections between global AI hubs to foster a shared understanding of critical issues and align on best practices.

Enterprise AI Steering Committee

A leading forum addressing the most important questions on the future of AI, our Enterprise AI Steering Committee defines and guides responsible AI use for enterprises and organizations.

New Partners

We welcomed 13 new global partners including nonprofits, research institutions, and private companies from around the world.

“Partnership on AI is committed to multidisciplinary approaches prioritising human well-being and ethical considerations. This partnership enhances our global reach and combines our complementary expertise to ensure AI technologies are developed and deployed to benefit society while minimizing potential harms.”

Prof. Sarvapali (Gopal) Ramchurn
CEO
Responsible AI UK

Preparing for the Age of Agents

As agents become more capable and widespread, so do their risks. Understanding agent capabilities, global governance challenges, and the right questions to ask are essential in preparing for their impact. This year, we focused on understanding these challenges in three timely reports.

Prioritizing Real-Time Failure Detection in AI Agents

When agents take actions, the nature of risk fundamentally changes. To address these risks, we need real-time failure detection.

AI Agents & Global Governance: Analyzing Foundational Legal, Policy, and Accountability Tools

Agents can amplify threats that cross national borders. Addressing these challenges requires more than national regulation, it requires global governance.

Preparing for AI Agent Governance: A Research Agenda for Policymakers and Researchers

A challenge for policymakers is not whether to regulate agents, but how when impacts remain uncertain. Understanding the evidence needed to make informed decisions is the first step.

“Ensuring the safety, security, and trustworthiness of AI is a global challenge that requires collective wisdom and cooperation. By joining the Partnership on AI, we look forward to contributing Korea’s perspectives and experiences while learning from our peers worldwide.”

Minn Seok Choi
Assistant Director for AI Safety Policy and Strategic Cooperation
Korea AISI

AI for Human Flourishing

As AI capabilities have evolved, we’ve collaborated with partners across industries to prepare for its potential impacts. From impacts to labor and workers, to trust in information and media, this year highlighted the growing need to understand how AI impacts people and society.

Workers Participating in Transparency: Addressing the Gap in AI Transparency on Labor

Transparency efforts largely omit impacts on workers. To address this gap, we must include workers as a topic, audience, and participants in transparency.

Safeguarding Trust and Dignity in the Age of AI-Generated Media

Based on the work from our analysis of case studies submitted by 18 organizational supporters of the Synthetic Media Framework, we crystallized key actionable takeaways.

From Deepfakes to Disclosure: PAI Framework Insights from Three Global Case Studies

The final round of case studies from Google, and civil society organizations Meedan and Code for Africa, in support of our Synthetic Media Framework.

Partnership on AI’s Newest Area of Work: AI and Human Connection

Built upon our work in AI and Media Integrity, we are ultimately responding to the pressing question: How can AI strengthen and sustain informed and connected communities?

“We are glad to deepen our collaboration with Partnership on AI by joining as a partner. PAI has long been a leader in addressing AI’s most important challenges, and we welcome the opportunity to continue to collaboratively work with the broader community.”

Dr. Rebecca Portnoff
Head of Data Science & AI
Thorn

Inclusivity and Trust in AI

AI is reshaping our lives and society. Inclusivity and trust in AI between people impacted by it and organizations using it remains a challenge. People impacted by AI are left out of critical conversations while the most vulnerable people are left to grapple with discrimination at a systemic level. In 2025, we narrowed in on how to bridge this gap.

Guidance for Inclusive AI

It is more important than ever for the people and organizations who develop and deploy AI-driven systems to work in close partnership with people impacted by it.

Participatory & Inclusive Demographic Data Guidelines

The rapid adoption and innovation of AI systems has left users vulnerable to algorithmic discrimination. To address these challenges, we developed the Participatory & Inclusive Demographic Data Guidelines.

Disclosure of AI-related Impacts, Risks, and Opportunities

As corporate AI use spreads, decision-makers need insight into how companies address issues such as bias, privacy, and environmental impacts.

Responsibly Navigating the Enterprise AI Landscape

How organizations use AI plays a critical role in shaping its real-world impact. Our latest report details the challenges and opportunities enterprises face in responsible AI adoption.

“Our collaboration within the Enterprise AI Steering Committee is vital. By creating a community of practice focused on the enterprise, we can guide the responsible development of AI to benefit society.”

Paula Goldman
Chief Ethical and Humane Use Officer
Salesforce

Shaping Global Policy

Rapid advancements in AI are presenting exciting opportunities but also significant challenges. In 2025, we focused on advancing global policy through documenting the impacts of foundation models, analyzing interoperability in governance, and informing policy strategy.

Documenting the Impacts of Foundation Models: A Progress Report on Post-Deployment Governance Practices

This report examines how model providers are implementing documentation, identifies challenges, and offers recommendations.

Decoding AI Governance: A Toolkit for Navigating Evolving Norms, Standards, and Rules

In AI governance, clarity and alignment is needed to avoid negative outcomes from fragmentation and geopolitical volatility.

Shaping the EU AI Act’s Code of Practice

The EU’s General-Purpose AI (GPAI) Code of Practice was approved by the AI Board and EU Commission, confirming that it meets the requirements of the AI Act.

“The governance choices we make today will determine how fully society can benefit from AI. By collaborating with PAI . . ., we are strengthening a global community committed to practical solutions, shared learning, and responsible innovation.”

Miriam Vogel
President and CEO
EqualAI

Who we are

PAI seeks to ensure that AI technologies are built to benefit and empower as many people as possible.

Our work is shaped by a global community of partners committed to the responsible development and use of AI. With the help of our PAI community we are able to advance positive outcomes for people and society.

Demographics

Staff

Board of Directors

Thanks to our Funders and Partners

PAI is an independent, nonprofit 501(c)(3) organization and funded by charitable contributions from philanthropy and corporate entities. Our 990 forms from previous years can be found on our funding page.

We thank all of our Partners who form this special community for their valuable insight and support.

A&O Shearman

ACLU

Ada Lovelace Institute

Adobe

AI4ALL

AI and Faith

AI Forum of New Zealand

AI Now Institute

Ai2 (Allen AI)

Alliance for AI

Amazon

American Association for the Advancement of Science (AAAS)

American Psychological Association

Annie E. Casey Foundation

Apple

Article 19

Associated Press

Association for Computing Machinery

Association for the Advancement of Artificial Intelligence (AAAI)

3A Institute (ANU School of Cybernetics)

BBC

Berggruen Institute

Berkeley Center for Law & Technology

Berkman Klein Center at Harvard University

Business for Social Responsibility

CBC/Radio-Canada

Capital One

Carnegie Endowment for International Peace

Carnegie-Tsinghua Center for Global Policy

Center for Data Innovation

Center for Democracy and Technology

Center for Human-Compatible AI

Center for Information Technology Policy

Centre for Internet & Society

Chatham House

Chief Executives for Corporate Purpose

CIFAR

Code for Africa

Concordia AI

Cornell CIS

Creative Commons

Credo AI

Data & Society

Data Economy Policy Hub

DataKind

Data Privacy Brazil

DeepMind

Demos

Digital Asia Hub

Digital Catapult

Duke Reporter’s Lab

EqualAI

Eticas Foundation

Fairly AI

Ford Foundation

Fraunhofer IAO

Full Fact

Future of Life Institute

Future of Privacy Forum

G3ict

GLAAD

Google

Google.org

Hacks/Hackers

Humane Intelligence

HCRI, Brown University

Human Rights Data Analysis Group

HRI Lab Tufts

IBM

IKEA – Ingka Group

Impact AI

Innovation Center for Artificial Intelligence (ICAI)

Insight SFI Research Centre for Data Analytics

Institute for the Future of Work

Intel

Internet Archive

Interpublic Group (IPG)

Intuit

Invisible Technologies

Joint Center for Political and Economic Studies

JPMorganChase

Korea AI Safety Institute

Leverhulme Centre for the Future of Intelligence (CFI)

Longpath Labs

MacArthur Foundation

Markkula Center for Applied Ethics

Mastercard

Meedan