Key Takeaways
- Anthropic launched the Anthropic Institute on March 11, 2026, to study AI’s economic, legal, and societal challenges
- Co-founder Jack Clark leads the Institute in a new role as Head of Public Benefit, after more than five years as Head of Public Policy
- The Institute launches with approximately 30 people and combines three existing Anthropic research teams
- Jack Clark believes powerful AI will arrive by end of 2026 or early 2027, driving urgency behind the Institute’s formation
Anthropic just formalized what many in AI feared was missing: a dedicated institution to track, study, and publicly report the real consequences that frontier AI could cause before it is too late. The Anthropic Institute is not a PR effort. It is a structured, interdisciplinary research body with a unique advantage: access to information that only the builders of frontier AI systems possess. What it does with that access over the next two years may shape how governments, workers, and industries respond to the most compressed period of technological change in modern history.
What the Anthropic Institute Actually Does
The Institute brings together three previously separate Anthropic research teams: the Frontier Red Team, which stress-tests AI systems to understand the outermost limits of their current capabilities; Societal Impacts, which studies how AI is being used in the real world; and Economic Research, which tracks AI’s effect on jobs and the broader economy. Together, these teams give the Institute a perspective no independent think tank can replicate because they have direct access to frontier models.
The Institute’s stated mission is to report what Anthropic learns while building powerful AI systems and to partner with external audiences to help address the risks those systems introduce. It will also engage with workers and industries facing displacement, using those interactions to inform both what the Institute studies and how Anthropic as a company chooses to act. This two-way structure is a meaningful departure from typical corporate research units.
Why Anthropic Is Building This Now
Anthropic’s own analysis signals that AI progress is compounding. Each capability breakthrough accelerates the next. In just five years, the company went from releasing its first commercial model to building systems that can discover severe cybersecurity vulnerabilities, take on a wide range of real work tasks, and begin to accelerate the pace of AI development itself.
Jack Clark told The Verge he believes powerful AI will arrive by the end of 2026 or early 2027. That conviction about the pace of progress is the direct reason he decided to shift from his policy role and build the Institute now rather than later. The pace of AI progress, in his words, is not slowing down for external events.
Jack Clark’s Role and the Context Behind It
Clark co-founded Anthropic and served as its Head of Public Policy for more than five years before this transition. The Institute’s launch also coincides with a significant moment for Anthropic: the company is currently in a legal dispute with the US Department of Defense after being designated a supply-chain risk, which would bar its clients from using Anthropic’s technology in their DoD work.
Clark has stated that the situation has not directly changed the Institute’s planned research agenda, but that it has affirmed the decision to release more information to the public. “What we’re experiencing with the last few weeks just sort of shows you how much hunger there is for a larger national conversation by the public about this technology,” he said.
The Founding Team
The Anthropic Institute launched with three named founding hires:
- Matt Botvinick, a Resident Fellow at Yale Law School and previously Senior Director of Research at Google DeepMind and Professor in Neural Computation at Princeton, leads the Institute’s work on AI and the rule of law
- Anton Korinek, Professor of Economics at the University of Virginia (on leave), leads research studying how transformative AI could reshape the very nature of economic activity
- Zoë Hitzig, who previously studied AI’s social and economic impacts at OpenAI, joins to connect the economics research to model training and development
Clark expects the Institute’s staff to double every year for the foreseeable future.
The Research Agenda: Four Areas That Matter
AI and the legal system. Matt Botvinick will lead a new team studying how powerful AI interacts with legal structures, including regulatory oversight and rule-of-law questions. This is among the least-studied risk areas in current AI governance literature.
Jobs and economic displacement. Anton Korinek and Zoë Hitzig will lead large economic research projects tracking how transformative AI could reshape economic activity and the labor market. Their work draws on Anthropic’s existing Economic Research team data.
Forecasting AI capabilities. A new forecasting team is being incubated within the Institute to predict how quickly AI capabilities will escalate. Accurate forecasts give policymakers lead time to build appropriate guardrails before deployment outpaces regulation.
Human and emotional impacts of AI. The Institute plans to conduct large-scale social science research into how AI use changes people over time, including studying emotional dependence on AI systems. Clark described the ambition as understanding “how the use of AI changes people” in the same way researchers studied the societal effects of social media.
Anthropic’s Expanding Policy Presence
Alongside the Institute, Anthropic is expanding its Public Policy team and opening its first Washington DC office this spring. Sarah Heck, who joined Anthropic as Head of External Affairs, will lead this team in her new role as Head of Public Policy. Before Anthropic, she served as Head of Entrepreneurship at Stripe and previously led global entrepreneurship and public diplomacy policy at the White House National Security Council.
The Institute and the Policy team operate as separate but complementary functions: the Institute generates evidence and reports findings publicly; the Policy team uses that evidence to engage with governance and regulation.
Limitations and Considerations
The Anthropic Institute is an internal body funded by and housed within Anthropic. Its ability to publish findings that are commercially inconvenient remains an open question. Clark acknowledged this directly, stating that Anthropic’s cofounders share similar values around public disclosure and that CEO Dario Amodei aligned on the importance of transparency even when it creates PR challenges. Independent verification by external researchers and regulators will be essential to establish full credibility over time.
Frequently Asked Questions (FAQs)
What is the Anthropic Institute?
The Anthropic Institute is an internal research body launched by Anthropic on March 11, 2026. It studies the societal, economic, and legal challenges that powerful AI will create. It combines three existing Anthropic research teams and is led by co-founder Jack Clark.
Who leads the Anthropic Institute?
Jack Clark, Anthropic’s co-founder, leads the Institute in his new role as Head of Public Benefit. He previously served as Anthropic’s Head of Public Policy for more than five years before transitioning into this new position.
What three teams make up the Anthropic Institute?
The Institute unifies the Frontier Red Team, which stress-tests AI systems at their capability limits; the Societal Impacts team, which studies real-world AI use; and the Economic Research team, which tracks AI’s effect on jobs and economies.
Why did Anthropic launch the Institute in 2026?
Anthropic believes AI progress is compounding and that extremely powerful AI is arriving sooner than most expect. Jack Clark stated he believes that threshold will be reached by end of 2026 or early 2027, making it urgent to study and report on AI’s societal impact now.
How large is the Anthropic Institute at launch?
The Institute launched with approximately 30 people, including founding members Matt Botvinick, Anton Korinek, and Zoë Hitzig. Jack Clark stated he expects the team’s headcount to double every year for the foreseeable future.
Is the Anthropic Institute independent from Anthropic?
No. The Institute operates within Anthropic and is funded by the company. Clark acknowledged the transparency challenge this creates but pointed to Anthropic’s public benefit corporation status and the shared values of its co-founders as structural commitments to candid reporting.
What is the connection between the Anthropic Institute and the Pentagon dispute?
The Institute launched days after Anthropic sued the US Department of Defense over a supply-chain risk designation. Clark said the dispute has not changed the research agenda but has reinforced the need to provide the public with more information about powerful AI.
