Anthropic

AI Safety · Research · San Francisco · London · Remote

AI safety lab building reliable, interpretable systems — and Claude.

4.5· 320 reviews

89%recommend92%CEO approval

Rating breakdown

320 reviews

4.5

Overall

People

1,300 people

San Francisco

Founded

2021

Industry

AI Safety · Research

AnthropicReview· 2mo

LALate SubmitVerified· 88 karma

Pros

Working with researchers and engineers at the top of their craft. Comp is competitive with FAANG, and the work feels important.

Cons

Pace is intense — easy to burn out if you don't set boundaries. Lots of context to absorb if you're not from a research background.

AnthropicInterview· 2mo

LALate Submit· 88 karma

Recruiter screen → live coding (React) → system design (frontend perf focus) → behavioural + values fit. Mission alignment is checked seriously.

Questions asked