
Researcher
Current role not listed
Sharma studied why AI assistants agree with users at the expense of truth.[1] He co-developed Constitutional Classifiers to block attempts to bypass AI safeguards.[2] He left Anthropic in February 2026, citing pressures to set aside core values and a desire to pursue writing and work more aligned with his integrity.[3]
AnthropicJul 2023-Feb 2026Work and education history is primarily focused on AI-relevant roles and may not be comprehensive.
First-listed author; assistants sacrifice truthfulness to match user beliefs.
Co-author of safeguards that defend against attempts to bypass model safety rules.
First-person resignation letter: February 9 departure, pressure on values within himself, Anthropic and society; desire for integrity and writing. Does not allege a specific triggering policy or dispute.
Based on public sources. If accuracy is important, verify details using the links or references.
For corrections, updates, or requests to remove information, contact us.