Jack Clark

Jack Clark co-founded Anthropic and, since March 2026, has been its Head of Public Benefit, leading The Anthropic Institute. The institute is Anthropic's internal think tank; it absorbed the Frontier Red Team, Societal Impacts and Economic Research teams and added work on forecasting AI progress and on how powerful AI systems will interact with the law. For Anthropic's first five years he ran its policy work.

He came to AI as a journalist. Georgetown's Center for Security and Emerging Technology describes him as having been "the world's only neural network reporter" at Bloomberg and the only distributed-systems reporter at The Register. In 2016 he joined OpenAI as director of strategy and communications, a title he still held in mid-2018, and later became its policy director. He left with Dario Amodei's group at the end of 2020.

Alongside his jobs he has built the field's public record. He writes Import AI, a weekly newsletter on AI research that his own site says reaches about 70,000 readers, was a founding member of Stanford's AI Index from 2017 to 2024, and sat on the first US National AI Advisory Committee. On 18 July 2023 he briefed the UN Security Council at its first meeting on AI, telling it that governments had to build ways to test AI systems.

In October 2025 he published "Technological Optimism and Appropriate Fear," comparing AI systems to shapes in a dark bedroom that people hope are "a pile of clothes on a chair." David Sacks, the White House AI adviser, responded that "Anthropic is running a sophisticated regulatory capture strategy based on fear-mongering."