Anthropic launches new think tank during Pentagon blacklist dispute

Anthropic, a US AI startup, launched the Anthropic Institute to study AI’s societal and economic impacts and named co-founder Jack Clark to lead it as head of public benefit.

Clark said he had “no concerns” about devoting resources to long-term research for the think tank despite potential short-term revenue impacts and expects powerful AI by late 2026 or early 2027.

The institute merges Anthropic’s societal impacts, frontier red team, and economic research groups and starts with about 30 staff including Matt Botvinick, Anton Korinek, and Zoe Hitzig.

Sarah Heck will lead public policy and the company plans a Washington, DC office focused on national security, AI infrastructure, energy, and democratic leadership in AI.

The announcement follows Anthropic’s lawsuit challenging a US government supply-chain risk designation that the company says would bar its clients from using its technology in their work with the Department of Defense.

🔗 Source: The Verge

🧠 Food for thought

Implications, context, and why it matters.

The institute’s launch follows a clash over limits on military and surveillance uses of AI

The lawsuit stems from Anthropic’s refusal to let its AI models support fully autonomous weapons and mass domestic surveillance of Americans 1.
The U.S. Department of Defense (DOD) demanded unrestricted access to the Claude model (Anthropic’s flagship AI assistant) for “all lawful purposes,” and Anthropic declined 2.
After that decision, President Trump ordered federal agencies to stop using Anthropic’s technology, and Defense Secretary Pete Hegseth told the DOD to label the company a “supply chain risk” 2.
That label is unusual because the authority was built to block foreign adversaries from damaging national security systems, rather than U.S. companies 3.

Anthropic’s legal fight raises questions about procurement power and speech

The dispute tests whether the government can use procurement law in ways Anthropic says punish the company for protected speech and its ethical policies 1.
Employees from OpenAI and Google backed Anthropic through an amicus brief (a “friend of the court” filing from non-parties) 2.
The brief called the government’s response an “improper and arbitrary use of power” 2.
A ruling could influence how AI companies bargain over limits on military uses of their technology 3.