Anthropic adds Novartis CEO to its board

Anthropic said its Long-Term Benefit Trust appointed Novartis CEO Vas Narasimhan to the AI company’s board, a move that gives trust-appointed directors a majority and adds healthcare experience to its governance.

Narasimhan is a physician-scientist and chief executive of Novartis.

Anthropic said the trust is an independent body with no financial stake that helps balance shareholder interests with the company’s public benefit mission.

He joins directors Dario Amodei, Daniela Amodei, Yasmin Razavi, Jay Kreps, Reed Hastings, and Chris Liddell.

Anthropic cited his experience in regulated industries and drug development.

🔗 Source: Anthropic

🧠 Food for thought

Implications, context, and why it matters.

Anthropic’s structure responds to governance turmoil at rival OpenAI

Anthropic is a Delaware Public Benefit Corporation, a legal status that lets its board weigh stockholder returns alongside its public benefit purpose of responsibly developing and maintaining advanced AI for the long-term benefit of humanity 1.
The setup aims to reduce pressure to ship an AI model before the company believes it is safe, a worry Anthropic leaders have raised when discussing the market for advanced AI systems 2.
Large backers such as Google and Amazon, which have invested billions, do not hold voting shares. Anthropic general counsel Brian Israel said they cannot elect board members or vote to change the trust’s governing rules 2.
Unlike OpenAI’s nonprofit-controlled structure, directors appointed by the trust still owe fiduciary duties to Anthropic’s stockholders, which narrows their ability to ignore financial interests 2.

This governance model now meets market pressure

Anthropic has called the Long-Term Benefit Trust (LTBT) a “corporate governance experiment.” It is meant to help the company balance public benefit with stockholder interests at turning points, rather than in daily decisions 1.
The arrangement has limits. A “failsafe” lets a supermajority of stockholders rewrite the rules that govern the LTBT without the consent of its five members, as described by TIME, the news magazine 2.
A strong result could offer a blueprint for other companies building powerful technologies with serious societal risks 1.
A weak result could bolster critics, cited in TIME, who argue that company governance alone falls short and that binding government regulation is needed for safe development of advanced AI 2.