Nathan Calvin on Frontier AI Safety and Governance

A sourced history of statements and actions associated with this topic.

Disputes over the development, testing, oversight, pacing, and governance of highly capable AI systems and their potential catastrophic risks.

Scope: Includes attributable disputes about whether frontier developers’ safeguards, deployment decisions, and proposed coordination mechanisms are sufficient. It does not treat speculative future outcomes as established facts.

Condemned1Distinct events · September 9, 2026 — Not selected; activate to add.Defended or excused0Distinct events · No published events — Not selected; activate to add.Challenged the characterization0Distinct events · No published events — Not selected; activate to add.Minimized0Distinct events · No published events — Not selected; activate to add.Mixed or conditional0Distinct events · No published events — Not selected; activate to add.Explicitly stayed out0Distinct events · No published events — Not selected; activate to add.

Counts document individual events. They do not assign an overall position.

By category Timeline

Distinct documented events

Position history

1 event on this page. Dates: September 9, 2026

1 matching event. Dates: September 9, 2026

1 total published event. Dates: September 9, 2026

September 9, 2026Condemned
Addressed record
Is Anthropic acting responsibly while advancing toward recursive self-improvement?
“Extremely powerful and urgent thread, please read.”
Speaker capacity
General Counsel at Encode, according to Encode's official biography; posted on his own account.
Editorial rationale
Endorses and amplifies Coxon's criticism; does not supply independent corroboration.
Nathan Calvin on Coxon’s warning and the frontier AI safety debate (opens in a new tab) · Nathan Calvin
Full primary public post and available surrounding thread. Related context: https://encodeai.org/team-members/nathan-calvin/ Speaker and position scope are explained in the linked event.

View context Is Anthropic acting responsibly while advancing toward recursive self-improvement?