How to Get an Anthropic Employee Referral
Anthropic is one of the most selective employers in AI, and demand for roles vastly exceeds openings. A warm introduction from someone the team trusts is close to essential.
Find Contacts Who Can Refer YouFree · No sign-up · See results in ~10 seconds
“Hi — we both studied at [your school]. I’d love to hear about your path to Anthropic before I apply…”
By the Numbers
Anthropic, the AI safety company behind Claude, receives an enormous volume of applications relative to its size. Hiring emphasizes both technical depth and genuine engagement with the company's mission of building reliable, safe AI systems — generic AI-hype applications stand out in the wrong way.
Referrals work the way they do at most startups-at-scale: employees flag candidates they believe in, and those applications get a real read. Because the company is mission-dense, a referrer who can speak to why you care about safe AI deployment — not just AI — adds signal a resume cannot.
How to Get a Referral: Step by Step
- Find a genuine connection: Use FindWarmIntros to surface Anthropic employees who share your school or a past employer — research labs, ML teams at big tech, and policy orgs are common backgrounds.
- Engage with the actual work: Read the team's published research or product documentation relevant to your target role. Reference it specifically; it is the clearest costly signal you can send.
- Match the role family: Research, engineering, product, policy, and go-to-market have different bars and different pipelines. Ask your contact which fits your background before the referral ask.
- Keep outreach short and concrete: Two or three sentences on who you are, one on why this team, one clear ask. Mission-driven companies get walls of text daily; brevity stands out.
- Apply to a specific posting: Anthropic lists roles publicly; have the exact role ready so your referrer can attach your name to a real requisition.
Tips That Make the Difference
Non-research roles are the wider door
Research scientist roles draw fierce competition. Product engineering, infrastructure, trust & safety, policy, and GTM teams have grown quickly and consider a broader range of backgrounds.
Public artifacts matter
Open-source work, evals, red-teaming write-ups, or thoughtful published analysis of model behavior demonstrate fit better than credentials alone.
Mission fluency is screened
Interviews probe how you think about AI risk and deployment tradeoffs. If your referrer can honestly say you engage seriously with those questions, their referral carries more weight.