Researcher · Founder

Andi Peng

AP

Current role not listed

Peng worked on reinforcement learning and post-training for Claude 3.5 through 4.5 at Anthropic. Her research asks how AI agents can learn nuanced human preferences from feedback, combining technical alignment work with experience in government technology policy.

See something inaccurate or outdated?

Background

Previously

2026-Jul 2026Co-founder,

Researcher,

About this data

Work and education history is primarily focused on AI-relevant roles and may not be comprehensive.

Last editorial review: August 8, 2026

News 2

moveAndi Peng leaves Humans&

The co-founder said July 30 was her last day at the human-centric AI lab.

July 30, 2026 person_addedAndi Peng added to the database

Humans& co-founder and former Anthropic researcher.