99 real questions reported from Anthropic interviews · updated Oct 2026.

Anthropic is an AI safety company building the Claude family of frontier models and enterprise AI products, pairing safety research with a rapidly growing product and API business. Interviews favor realistic engineering work and judgment over puzzle-style algorithms.
Anthropic is an AI safety company building the Claude family of frontier models and enterprise AI products, pairing safety research with a rapidly growing product and API business. Interviews favor realistic engineering work and judgment over puzzle-style algorithms.
Covered in the research, each with its source
Computed from candidates’ write-ups and our Anthropic questions.
Reported outcomes
3%
of the 68 decided loops shared here ended in an offer.
Outcome mix
Across the 198 of 198 shared loops that stated an outcome.
Self-reported by candidates who shared a Anthropic write-up. It describes those loops, not an application-to-offer rate. Read them.
What they ask
Topic mix
Across the 34 of 99 Anthropic questions tagged with a topic.
Difficulty mix
Across the 78 of 99 questions with a difficulty rating.

Anthropic
Anthropic
Anthropic
Looking for another company?
Browse all companies