AI

I believe the current moment in AI marks a turning point for humanity. We are scaling our ability to reason and interact with information in an unprecedented manner. At the very least this moment introduces a new set of transformative tools that will unlock strange and different futures. It's also possible that AI does more - that it supersedes our own agency. Such a future needn't come from AGI - the infrastructure of automated decision making systems we are creating are optimized for objectives beyond our direct control, often because we have limited understanding of the actual goals, behaviors, and emergent consequences of the systems we create.

In light of the potential and risks of AI, I have turned my career towards improving our "wisdom" in deploying AI systems. I do this by fostering community discussion about the impact of AI via the Ai Salon and by maturing AI Governance practices. The Ai Salon has grown into a 7,000-member global community that has hosted 100+ in-person events across San Francisco, Bangalore, Vancouver, Lagos, London, and Berlin. I distill these conversations on our substack for broader benefit, and build the AI tooling that helps chapter leads run events and synthesize what emerges.

I broadly see AI Governance developing the organizational capacities, institutional wisdom, and regulatory frameworks to properly incentivize and manage AI development. I work on this as Head of AI & Research at Credo AI, where I lead the teams and systems behind our AI-native governance products — including Agent Governor, which governs AI agents at the harness layer, and the Unified Control Framework, a common foundation connecting risks, controls, and regulations across global frameworks.

I have written on a number of topics relating to AI governance, and have contributed to policy-relevant RFCs (e.g., the NIST AI RMF), academic articles like the Unified Control Framework and the MLCommons AI Safety Benchmark, and blogs for the general public including: what AI governance is, how to operationalize it, and how governance applies to Generative AI systems.

I've also been privileged to contribute to a number of multistakeholder AI Safety initiatives, including the US AISIC, MLCommons AI Safety working group, Partnership on AI working groups, and NIST's AI risk management framework, both through direct commentary in collaboration with Assistant Secretary Alan Davidson and Credo AI, by contributing to other projects, and through panels as part of NIST's RMF workshops. More recently I contributed to the Partnership on AI's 2026 Transparency Report on Foundation Model Impacts and the MIT AI Risk Initiative, and joined the founding Exploratory Committee of Fathom, an AI governance nonprofit.