Anthropic names ex-justice Cuéllar as global affairs chief

The interesting bet here is not that Anthropic hired a policy person; every frontier lab has one. It is who the policy person is. Mariano-Florentino Cuéllar, an ex-California Supreme Court justice who ran the Carnegie Endowment for International Peace until July and served as a special assistant in the Obama White House, is now Anthropic's first chief global affairs officer, reporting to president Daniela Amodei out of San Francisco, TheNextWeb reports.

The context is what makes the pick pointed. Earlier this year the Pentagon blacklisted Anthropic's technology after a standoff over military red lines, and Anthropic is challenging that designation in court. Separately, the Trump administration handed down an export control directive that temporarily banned Anthropic from selling its top-shelf Mythos 5 and Fable 5 models to foreign buyers on national security grounds. Hiring a Democratic-connected national-security lawyer as your ambassador to a Republican White House is either brave or a signal that the fight matters more than the optics.

Cuéllar comes with genuinely on-thesis credentials. He co-chaired a Carnegie task force on US national security and nuclear proliferation, which is the historical analogue policymakers keep reaching for on AI dual-use, and a 2025 study he co-led helped inform California's SB 53. That law protects AI whistleblowers, requires high-revenue developers to report on safety standards and incidents, and carries a maximum $1 million fine per violation. Anthropic endorsed the bill. So the company is now hiring the person whose research helped shape rules it publicly asked for.

There is a governance wrinkle worth flagging. Cuéllar has served on Anthropic's Long-Term Benefit Trust since January, the independent body that exists to keep the public-benefit mission honest, and he is stepping down from that seat to take an operating role. What the reporting from The Star does not give you is a named successor to that trust seat, or what Anthropic committed to in order to get the export ban lifted, so treat the standoff-and-repair arc as reported, not as settled.

The upside for Anthropic, if this works, is not just a truce with Washington. It is authorship. If Cuéllar helps shape the emerging US AI-safety regime from inside a frontier lab, the rules the rest of the industry ends up living under will look a lot like the ones Anthropic already said it wanted.