On March 11, Anthropic unveiled the Anthropic Institute — a new internal research body dedicated entirely to studying the societal impact of AI. Leading the charge is co-founder Jack Clark, who’s stepping down from his role as Head of Public Policy to take on a new title: Head of Public Benefit.
Three Teams Become One
The institute brings together three existing Anthropic teams: the Frontier Red Team (which stress-tests AI systems at their limits), the Societal Impacts group, and the Economic Research team. They’ve also brought in some serious talent: Matt Botvinick will research AI and legal systems, economist Anton Korinek will study how transformative AI reshapes economic activity, and Zoe Hitzig will bridge economic insights with AI development.
Why Now?
The reasoning is remarkably candid. Anthropic writes that they expect ‘far more dramatic progress in the next two years’ and that extremely powerful AI is arriving ‘far sooner than many think.’
That’s not marketing fluff. When a company that builds frontier models establishes its own think tank to study the consequences of its own technology, it signals something: they’re not just building — they’re thinking about what they’re building.
More Than Just Research
In parallel, Anthropic is expanding its Washington presence. Sarah Heck has been named Head of Public Policy, and a D.C. office is set to open this spring. Given the ongoing battle with the Pentagon, this is a strategic move — Anthropic wants to fight not just in court but in the policy arena.
My Take
The Anthropic Institute is one of the most interesting moves any AI company has made so far. Instead of leaving the societal consequences of AI to others — politicians, journalists, activists — Anthropic is deploying its own economists, social scientists, and ML engineers to research these questions systematically.
Sure, you can read this as a PR play. But the personnel decisions tell a different story: Jack Clark isn’t just anyone — he’s a co-founder. And the researchers being hired have serious academic credentials.
Whether the institute can truly work independently while sitting inside Anthropic remains to be seen. But this approach beats the alternative: not looking at all.
Sources: