Amanda Askell

Amanda Askell is a philosopher from Scotland with a PhD from New York University, where she wrote her dissertation on infinite ethics. She moved from academia into AI at OpenAI, co-authoring the GPT-3 paper on its policy side, and in 2021 joined Anthropic in its founding cohort — one of the few people shaping a frontier lab whose training is in moral philosophy rather than machine learning.

The Claude whisperer

Askell leads the team responsible for Claude’s character and personality — colleagues call her the “Claude whisperer” — deciding through post-training what disposition the model should have. Her template is “a well-liked traveler who can adjust to local customs and the person they’re talking to without pandering to them”: honest, curious, and unwilling to simply capture attention or flatter. She is the primary author of Claude’s constitution, the document that articulates the model’s values, and named to the TIME100 AI list in 2024.

Her work reframes alignment as character formation rather than rule enforcement — teaching a system judgment instead of compiling prohibitions. The Wall Street Journal put it plainly in 2026: her job is to teach Claude how to be good. Askell has also warned against AI presenting itself as “a grand source of authority,” arguing users should trust model outputs critically, not reverently.