Anthropic's product manager job descriptions pay a stated base of $305,000 to $460,000 and ask far more for ambiguity, research partnership, safety judgment and evals than for SQL or A/B tests. We read all 19 Anthropic PM postings in the AllthingsPM jobs catalog (read from Anthropic's Greenhouse board in September 2026). Ten list $305,000 to $385,000, eight list $385,000 to $460,000, and one spans both. Every role sits in San Francisco, New York, Seattle or Washington, DC. Comfort with ambiguity appears in 15 of 19, research teams in 14, safety and risk judgment in 14 and evals in 10. SQL appears in 3 and A/B testing in 1. On 7 November 2026 the live board listed 16 PM roles out of 675 jobs, in the same two pay bands.
How did we analyse Anthropic's PM job descriptions?
We pulled every Anthropic posting from the AllthingsPM jobs catalog on 25 September 2026; the catalog had read them from Anthropic's Greenhouse board on 22 September. We read each body in full, set aside the boilerplate all 19 share, and hand-coded each for team, locations, years, salary and 14 requirement themes. Two roles (Growth and Beneficial Deployments) had already come down by 25 September but were posted that month, so we kept them. For the market comparison we ran the same text matches over 370 other PM postings from 85 AI companies in our job-description harvest.
What PM roles is Anthropic hiring for?
Here are all 19 from the September set, with the salary and experience each posting states. Every title links to its page in our jobs catalog, with the full posting and a mock interview built from it.
| Role | Team family | Locations listed | Annual salary stated | Years named (where) |
|---|---|---|---|---|
| Safeguards (Generalist) | Safeguards and trust | SF, NYC | $385,000 to $460,000 | 5+ in PM (preferred) |
| Safeguards (Account Integrity & Abuse) | Safeguards and trust | SF, NYC | $385,000 to $460,000 | 5+ in PM (preferred) |
| Safeguards (Cyber) | Safeguards and trust | SF | $305,000 to $385,000 | 5+ in PM (preferred) |
| Safeguards Rare Harms | Safeguards and trust | SF | $305,000 to $385,000 | 5+ in PM (preferred) |
| Safe Access | Safeguards and trust | SF, NYC | $305,000 to $385,000 | "five or more years" typical (minimum); 8+ preferred |
| Multi-Cloud Trust & Safety | Safeguards and trust | SF, NYC, Seattle | $305,000 to $385,000 | 10+ overall, 5+ in platform product (minimum) |
| Research | Research, Labs, data | SF, NYC | $305,000 to $385,000 | 5+ in PM (preferred) |
| Research (Code) | Research, Labs, data | SF, NYC | $305,000 to $385,000 | 5+ in PM (preferred) |
| Research Product Manager, Labs | Research, Labs, data | SF, NYC | $385,000 to $460,000 | 5+ in PM (preferred) |
| Beneficial Deployments (Labs) | Research, Labs, data | SF, NYC | $385,000 to $460,000 | 5+ in PM or operating roles (preferred) |
| Human Data Platform | Research, Labs, data | SF, NYC | $305,000 to $385,000 | 5+ in PM ("ideal candidate" list) |
| Claude Science | Claude products and growth | SF, NYC, Seattle | $305,000 to $385,000 | 8+ (preferred) |
| Claude Tag | Claude products and growth | SF, NYC | $385,000 to $460,000 | 8+ combined ("good fit" list) |
| Cybersecurity | Claude products and growth | SF, NYC | $385,000 to $460,000 | 12+ (preferred) |
| Growth | Claude products and growth | SF, NYC, Seattle | $385,000 to $460,000 | 6 to 10 in PM (preferred) |
| Multi-Cloud Growth, Google | Platform, cloud, public sector | SF | $305,000 to $460,000 | 10+ in developer software, 5+ in PM (preferred) |
| New Markets and Monetization | Platform, cloud, public sector | SF | $305,000 to $385,000 | "five or more years" typical (minimum); 8+ preferred |
| Public Sector | Platform, cloud, public sector | Remote-friendly (travel required), Washington, DC | $305,000 to $385,000 | roughly 7+ ("strong candidates") |
| Business Technology | Internal tools | Remote-friendly (travel required), SF, Seattle, NYC | $385,000 to $460,000 | None named |
Salary, location and years are as each posting states them. The team families are our grouping, not Anthropic's.
What changed on the live board by November 2026?
On 7 November 2026 Anthropic's Greenhouse board listed 16 PM postings out of 675 jobs. By title, 13 of our 19 are still there: Research, Research PM Labs, Claude Science, Cybersecurity, Multi-Cloud Growth, Multi-Cloud Trust & Safety, New Markets and Monetization, Public Sector, Safe Access and four Safeguards roles. Three titles are new: Claude Marketplace, Enterprise Privacy and Product Support Experience. Research (Code), Beneficial Deployments, Human Data Platform, Claude Tag, Growth and Business Technology no longer appear.
The pay bands have not moved. Most roles list $305,000 to $385,000; Cybersecurity, Safeguards Account Integrity, Safeguards Generalist, Product Support Experience and Research PM Labs list $385,000 to $460,000; Multi-Cloud Growth still spans $305,000 to $460,000. Roles open and close weekly, so check the live board before you apply.
Which Anthropic teams are hiring the most product managers?
Safeguards, trust and access (6). The four Safeguards postings describe a PM who works "with our research and product teams to develop detections, evals, interventions, and tools to measure and mitigate deployment and user risks". Safe Access and Multi-Cloud Trust & Safety are platform roles with a safety core. A practice prompt from our bank: a new Claude capability unlocks major user value but creates a rare, high-severity misuse risk.
Research, Labs and human data (5). The two Research postings are the most technical ("working proficiency in Python and SQL"). The two Labs roles are 0-to-1 and ask you to "build prototypes yourself to validate ideas before committing resources". Human Data Platform builds tooling for the data that trains Claude; we turned it into a question: design a KPI framework for Anthropic's Human Data Platform.
Claude products and growth (4). Claude Science wants people who can "hold their own in a conversation with a structural biologist or a computational chemist". Growth was the only posting that named A/B testing.
Platform, cloud and public sector (3), and internal tools (1). The multi-cloud and New Markets roles want someone who has owned "a revenue number, a margin target, or a channel P&L". Public Sector requires that you "are a US person". Business Technology asked you to "Design for where models will be in 3,6,12 months".
What does an Anthropic product manager get paid, and where?
Every posting gives an "Annual Salary" in USD. These are base salary ranges required by California and New York pay transparency law, not total compensation. Identical Safeguards texts sit in different bands, and every posting says "Years of experience required will correlate with the internal job level requirements for the position", so our reading is that the band reflects an internal level. That is an inference; ask your recruiter.
San Francisco appears on 18 postings, New York on 14 and Seattle on 4. Only Public Sector and Business Technology were remote-friendly, both with travel. The shared text expects staff "in one of our offices at least 25% of the time". No PM role is based outside the US. On visas, Anthropic says it does sponsor them but cannot do so for every role and candidate.
How senior do you need to be?
No title says Senior, Staff or Principal. Seniority shows in years and pay instead: 9 postings name 5+ years as their highest figure, 4 name 8+, 2 name 10+, Cybersecurity names 12+, and Business Technology named none, hiring "on demonstrated impact". In 13 postings the years are only preferred, and Safe Access and New Markets say "we care about the ownership, not the years". Every posting adds that you should apply even if you do not meet every qualification. There is no entry-level PM role in the set.
Because the postings weigh ownership over years, your resume has to prove ownership. Our JD resume review checks your resume against the exact Anthropic posting and flags requirements it does not yet evidence.
What skills do Anthropic PM postings ask for?
Evals are a PM skill here. Safeguards PMs write safety evals, Claude Science PMs "build and shape evals grounded in real scientific workflows", and Cybersecurity and Public Sector want "Fluency with AI products: evaluations, model behavior as part of the product surface". If you have never designed an eval set or a launch gate, close that gap first; the Designing Evals for AI Products chapter of our course covers it.
Research is your main partner. Fourteen postings name research teams, and Human Data listed "Researcher EQ" as a virtue. The lesson on working with researchers covers this, which general PM courses usually skip.
Minimum qualifications are mostly judgment and ownership; preferred ones carry the credentials. A few roles make domain depth a hard minimum, such as security for Cybersecurity.
How do Anthropic's PM postings compare with other AI companies?
| Term in the posting text | Anthropic (19) | Other AI companies (370) |
|---|---|---|
| "Ambiguity" or "ambiguous" | 79% | 51% |
| "Eval", "evals" or "evaluations" | 58% | 15% |
| 0-to-1, in any spelling | 53% | 11% |
| "Founder" | 37% | 16% |
| SQL | 16% | 10% |
| A/B testing | 5% | 4% |
All 19 Anthropic postings state a salary range, against roughly half of the others (46% to 52%, depending on how strictly you match). The practical point: a prep plan tuned to metrics and A/B tests misses what 10 of 19 postings ask for. For the wider market, see our state of AI PM hiring in 2026.
What does Anthropic say about its interviews?
Anthropic publishes a few rules but not a stage count or timeline. All interviews run on Google Meet. Non-technical interviews are conversational and focus on how you think and why you want the role. Anthropic does not give feedback, and you can reapply after 12 months. Its candidate guidance says not to use AI in live interviews or take-homes unless told otherwise, though AI is fine for preparing and refining your application. Round-by-round loops you see online come from third-party guides, not Anthropic.
How should you prepare for an Anthropic PM role?
- Practise on the posting's own text. Each role in the table has a scored mock interview in AllthingsPM built from its job description, in text or voice, with follow-ups. Our Anthropic company page adds 105 Anthropic questions, 90 written from these postings, each with an answer guide.
- Build an eval story. What you measured, how you built the test set, what launch decision it changed. The evals and Defining AI Guardrails and Agent Security chapters cover the PM side.
- Build something yourself with an AI tool. A small working prototype beats a claim; the Prototyping AI Products With Coding Tools chapter shows how to scope one.
- Prepare a safety tradeoff and a research partnership story. Practise with turning vague or contradictory technical feedback into a roadmap decision.
- Tailor your resume to the posting. Run it against the exact job description in the JD resume review, or use Resume Job Match to see which roles fit.
Where can you practise on these postings?
| AllthingsPM | Anthropic careers board | Generic AI chat | |
|---|---|---|---|
| All 19 September postings, with salary | Yes | Live roles only | No |
| Mock built from the posting | Text or voice, scored | No | If you set it up |
| Anthropic questions with answer guides | 105 | No | No |
| AI PM course (evals, trust) | 14 chapters, from 604 postings | No | No |
| Resume checked against the JD | Yes | No | If you set it up |
Human coaches add insider judgment. For daily practice, AllthingsPM is $12 a month list, with Chapter 1 of the course free and a 7-day free trial.
Frequently asked questions
How many product manager roles is Anthropic hiring for?
On 7 November 2026 Anthropic's Greenhouse board listed 16 PM postings out of 675 jobs. Our September analysis covers 19 postings, 13 of which are still listed by title. Safeguards remains the largest group, with four roles.
What is the salary for a product manager at Anthropic?
Posted base pay falls in two bands: $305,000 to $385,000 and $385,000 to $460,000, with Multi-Cloud Growth spanning $305,000 to $460,000. These are base ranges required by California and New York pay transparency laws, so they exclude equity and bonuses.
Do Anthropic PM roles require a technical background?
Most ask for technical depth, few ask for coding. Only the two Research postings named Python, and three named SQL. Claude Tag said "An engineering background is not required". The postings want enough depth to make technical tradeoffs, work with researchers and reason about evals.
How many years of experience does Anthropic want for PM roles?
The lowest figure is 5+ years and the highest 12+ (Cybersecurity). In 13 of 19 postings the years are only preferred, and two say they care about "the ownership, not the years".
Can you work remotely as a PM at Anthropic?
Mostly not. Staff are expected in an office at least 25% of the time. In our set only Public Sector and Business Technology were remote-friendly, both with travel. Public Sector requires a US person.
Sources
- AllthingsPM jobs catalog, 19 Anthropic PM postings (catalog_jds, read from Anthropic's Greenhouse board on 22 September 2026, analysed 25 September 2026): jobs
- Anthropic careers board on Greenhouse, live job list checked 25 September 2026 (624 jobs, 17 PM roles) and 7 November 2026 (675 jobs, 16 PM roles): job-boards.greenhouse.io/anthropic and boards-api.greenhouse.io/v1/boards/anthropic/jobs
- Anthropic, Product Manager, Safeguards (Generalist): greenhouse
- Anthropic, Product Manager, Safeguards (Account Integrity & Abuse): greenhouse
- Anthropic, Product Manager, Safeguards (Cyber): greenhouse
- Anthropic, Product Manager, Safeguards Rare Harms: greenhouse
- Anthropic, Product Manager, Safe Access: greenhouse
- Anthropic, Product Manager, Multi-Cloud Trust & Safety: greenhouse
- Anthropic, Product Management, Research: greenhouse
- Anthropic, Product Manager, Research (Code): greenhouse
- Anthropic, Research Product Manager, Labs: greenhouse
- Anthropic, Product Manager, Beneficial Deployments (Labs), no longer on the board as of 25 September 2026: greenhouse
- Anthropic, Product Management, Human Data Platform: greenhouse
- Anthropic, Product Manager, Claude Science: greenhouse
- Anthropic, Product Manager, Claude Tag: greenhouse
- Anthropic, Product Manager, Cybersecurity: greenhouse
- Anthropic, Product Manager, Growth, no longer on the board as of 25 September 2026: greenhouse
- Anthropic, Product Manager, Multi-Cloud Growth, Google: greenhouse
- Anthropic, Product Manager, New Markets and Monetization: greenhouse
- Anthropic, Product Manager, Public Sector: greenhouse
- Anthropic, Product Manager, Business Technology: greenhouse
- AllthingsPM JD corpus, 389 PM postings from 86 AI companies (370 non-Anthropic postings from 85 companies used for comparison), read 22 September 2026
- Anthropic, careers page and guidance on candidates' AI usage: anthropic.com/careers and anthropic.com/candidate-ai-guidance
- California Employment Law Report, "What California Law Requires in Your Job Postings", April 2026 (Labor Code 432.3, SB 1162, SB 642): californiaemploymentlawreport.com
- New York State Department of Labor, Pay Transparency (Labor Law 194-B): dol.ny.gov
- AllthingsPM question bank, Anthropic company page (105 questions, 90 written from Anthropic PM postings): question-bank/companies/anthropic
