Who Jack Clark is
Jack Clark is a co-founder of Anthropic — one of the handful of companies at the very frontier of AI development — and the author of Import AI, a weekly newsletter that has become a go-to digest for anyone trying to keep up with what's actually happening in AI research. He occupies an unusual position: he helps build some of the most powerful AI systems in the world while also being one of the field's most consistent voices about the risks those systems carry.
Why he matters
Most people in AI either build things or worry about them. Clark does both, publicly, and has done so for years. That combination gives him a perspective that's rare: he understands the technical details well enough to explain them clearly, and he takes the societal stakes seriously enough to keep asking uncomfortable questions — about alignment, about economic disruption, about what happens when AI systems start improving themselves.
Import AI: a weekly map of the frontier
Import AI is Clark's newsletter, published weekly and now past issue 464. Each issue synthesizes a handful of significant AI developments — research papers, capability milestones, policy moves, geopolitical signals — and frames them with Clark's own analysis. Recent issues have covered topics like:
- Whether AI alignment efforts are on track (issue 461)
- AI systems beginning to automate AI research itself — a step toward recursive self-improvement (issue 455)
- Scaling laws for cyberattacks and the industrialization of cyber espionage (issues 450, 452, 442)
- A reported 10,000-GPU Chinese compute cluster and what it signals about the global AI race (issue 463)
- Paths to artificial superintelligence and the quasi-religious nature of singularity beliefs (issue 462)
The newsletter doesn't just report — it asks questions. Issues regularly end with a framing provocation: when will financial markets start pricing in transformative AI? Is superintelligence a phase change or a gradual shift? How should we think about AI systems that persist through time?
The Anthropic Institute
In March 2026, Clark took on a new formal role at Anthropic: Head of Public Benefit, leading the newly established Anthropic Institute for AI Societal Impact Research. The Institute brings together three existing Anthropic teams — the Frontier Red Team, Societal Impacts, and Economic Research — under one roof, with a mandate to study how AI is reshaping economies, jobs, governance, and legal systems.
Founding researchers include Matt Botvinick (AI and the rule of law), Anton Korinek (transformative AI economics), and Zoë Hitzig (AI's social and economic impacts). The Institute launched alongside an expansion of Anthropic's public policy work and the opening of a Washington DC office — a signal that Anthropic is investing seriously in the policy conversation, not just the technical one.
The through-line
Whether writing Import AI or running the Anthropic Institute, Clark's consistent preoccupation is the gap between how fast AI is moving and how prepared the world is for what's coming. His newsletter covers the technical developments; his institutional role is about building the research infrastructure to understand — and ideally shape — what those developments mean for everyone else.
