A New Chapter: Reflecting on Humane Intelligence’s Transformative Impact

After an incredible journey building Humane Intelligence from the ground up, it’s bittersweet to share that I’ll be transitioning from day-to-day leadership of this remarkable organization. This moment calls for celebration—of what we’ve accomplished together and the foundation we’ve built for the future.

How far we’ve come

When we founded Humane Intelligence, we set out to democratize AI evaluation and build a community of practice around algorithmic assessment. Today, I’m proud to say we’ve made headway towards that vision and exceeded it in ways that continue to surprise and inspire me.

Most people learned of Humane Intelligence through our work running the largest-ever generative AI red teaming exercise at DEFCON 31, where 2,244 participants evaluated eight major language models across 21 critical topics. This groundbreaking initiative, developed with industry partners including Anthropic, Google, Meta, and OpenAI, and sponsored by the White House, demonstrated how public red teaming can complement corporate safety efforts and policy imperatives while bringing diverse perspectives to AI oversight. The transparency report and open-source dataset continue to inform research worldwide, proving that democratized evaluation isn’t just possible—it’s essential.

Our algorithmic bias bounty programs have tackled real-world challenges from counterterrorism to environmental justice. From detecting hateful imagery manipulation to supporting reforestation efforts in partnership with the Indian Forest Service, these competitions have proven that diverse communities can solve complex technical problems when given the right platform and incentives.

Beyond technical achievements, we’ve fostered international collaboration through events like AI & Society House in Paris, bringing together government officials, researchers, and civil society organizations from across the globe. Our partnerships span from UNESCO and the UK Royal Society to Singapore’s IMDA, demonstrating that responsible AI development requires global cooperation.

What’s next for me: Infrastructure for AI Evals

Looking ahead, I’m excited to announce the launch of a mission-aligned for-profit venture that will complement and amplify Humane Intelligence’s work. This new public benefit corporation will be developing and creating sustainable pathways for organizations to implement robust AI evaluation practices at scale. 

Humane Intelligence, the nonprofit will continue thriving under capable leadership, making our community stronger than ever. Our Board of Directors has endorsed a thoughtful transition plan, and our committed team is ready to expand the impact we’ve only begun to realize.

To everyone who contributed to this journey—our community members, partners, participants, funders, and supporters—thank you for proving that inclusive, rigorous AI evaluation isn’t just an ideal, but an achievable reality. The future of responsible AI depends on the foundation we’ve built together.