Dario Amodei

Overview

Dario Amodei is the co-founder and CEO of Anthropic. His public essays and policy posts have repeatedly anchored the corpus’s frontier-safety and frontier-policy threads — from the “cone of uncertainty” framing on multi-year data-center commitments to the recursive-self-improvement safety series the Anthropic Institute has been running.

Timeline

2026-06-12-AI-Digest — Amodei publishes a new essay calling for mandatory third-party frontier-model audits across four risk areas, government authority to block models in an “aviation-style” pre-deployment regime, compute-threshold disclosure (10^25 FLOP / $500M revenue / $1B R&D triggers, 15-day incident reporting), tiered labour-displacement support, and democratic-coalition chip-and-model export controls. The Decoder’s “Cold War playbook” framing is editorial gloss — Amodei reaches for nuclear-weapons analogies and national-security categories, not Cold War ones. Reads as ground-laying for an export-controlled, third-party-audited, government-block-authority regime that the lab’s own commercial posture already partly implements; whether the public-policy surface follows is a different question.
2026-06-15-AI-Digest — Amodei attends the G7 opening in Évian-les-Bains today (June 15–17) alongside OpenAI‘s Sam Altman and DeepMind‘s Demis Hassabis at President Macron’s personal invitation — the first time the three Western frontier-lab heads have collectively appeared before G7 governments. European labs (Mistral’s Mensch, Cohere’s Gomez, Stability’s Rombach) surface in the same coverage. Bloomberg frames the agenda around a voluntary commitments package on youth safety and AI-infrastructure coordination; the corpus’s disciplined read holds that the “industry-to-state bargaining shift” framing is editorial more than reportorial — what’s actually load-bearing is the structural fact that the joint appearance lands 48 hours after the Claude Fable 5 / Claude Mythos 5 global disable, with the export-control story, the IPO clock running on Anthropic and OpenAI, and the G7 voluntary-commitments framework now visibly running in the same negotiating window. Stacks against Amodei’s 2026-06-12 essay-as-policy-stack (2026-06-12-AI-Digest) as the same regulatory-regime vocabulary he was reaching for, now in the room with the G7.
2026-06-19-AI-Digest — Amodei is the named recipient of US Commerce Secretary Lutnick’s letter that restricted broader foreign access to Claude Fable 5 and Claude Mythos 5 starting June 12 — today’s Bloomberg confirmation that the Project Glasswing preview cohort retained access narrows the directive’s operational scope. The carve-out resolves a real ambiguity that had hung over the previewing partners since the public limits landed: Mythos 5 was the lab’s most aggressive vulnerability-discovery model and the Commerce action was the first documented enforcement of US export-control authority against a deployed commercial frontier model rather than against weights-at-rest or chips. The first carve-out inside any of the three 2026 export-control instruments touching frontier models lands on the regime Amodei’s June 12 essay was sketching, with the lab’s own posture and the federal directive now visibly engaged at the same operational layer.

Key Developments

Third-Party Frontier Audits + Compute-Disclosure Regime (June 12, 2026): First fully-formed public-policy stack from the Anthropic CEO of the year — mandatory third-party frontier-model audits across four risk areas, government pre-deployment block authority, compute-threshold disclosure tied to 10^25 FLOP / $500M revenue / $1B R&D triggers with 15-day incident reporting, tiered labour-displacement support, and democratic-coalition chip-and-model export controls. Reads alongside Project Glasswing‘s expanded US-government deployment of Claude Mythos 5 and the same-day Fable 5 apology as Amodei sketching the regulatory regime the lab’s own posture already implements.

Dario Amodei

Overview

Timeline

Key Developments

Related