AI product manager jobs are concentrated, senior and well paid. On 22 September 2026, AllthingsPM read the public job boards of 160 companies building AI products and found 389 open product manager postings at 86 companies. Databricks (20), Anthropic (19), Sierra (18) and Datadog (18) posted the most. Only 3 of the 389 titles were associate or APM roles. Posted pay bands at AI companies ran from about $160,000 to $460,000 in base salary.
AllthingsPM is an AI PM course and PM interview prep platform. The advantage for a job seeker is that we do not stop at the report: 116 of these AI company postings are live in the AllthingsPM jobs catalog, and each one opens a mock interview built from that exact job description.
This is the first edition of a quarterly series. Every edition uses the same seven sections, in the same order, so you can compare quarters directly.
What does the AI product manager job market look like this quarter?
Here is the quarter in one table. Next edition, the same rows will get a second column so the change is visible at a glance.
| Metric, Q3 2026 | Value | Source |
|---|---|---|
| AllthingsPM live AI company JDs with a mock | 116 at 18 companies | AllthingsPM/jobs |
| Company job boards read | 200 attempted, 160 reached | AllthingsPM JD corpus |
| Open PM postings found | 389 at 86 companies | AllthingsPM JD corpus |
| Postings clearly about AI products (tagged) | 335 at 88 companies | AllthingsPM demand profile |
| Posting mentions San Francisco | 141 of 389 (36%) | AllthingsPM JD corpus |
| Posting mentions New York | 90 of 389 (23%) | AllthingsPM JD corpus |
| Posting mentions remote | 67 of 389 (17%) | AllthingsPM JD corpus |
| Associate or APM titles | 3 of 389 | AllthingsPM JD corpus |
| Base pay band range seen | $160,000 to $460,000 | Salary text on company job boards |
Corpus read 22 September 2026 from Greenhouse (240 postings), Ashby (145) and Lever (4) job boards. The tagged profile is a separate pass over AI-native postings, so its counts differ slightly.
A note on method, because every hiring report should state one. We read company job boards directly, not aggregators, so every posting is a real, current opening with a full description. The trade-off is scope: this covers AI-native companies and AI teams at tech companies, not every "AI product manager" title in the market. For the wider picture, Axial Search tracked 12,397 US AI product postings in 2026 and found 93% senior level or above and 2% junior [1], and Lenny Rachitsky, using TrueUp data, counted over 7,300 open PM roles at tech companies globally in March 2026, 75% above the early 2023 low [2]. Both line up with what we saw: more openings, almost none of them junior.
How AllthingsPM does this. The same corpus powers the product. The AllthingsPM jobs catalog keeps 116 live AI company JDs, and each one opens a scored mock interview in text or voice with follow-ups. If you do not have a target role yet, Resume Job Match lists open roles and ranks them against your resume.
Which companies are hiring the most AI product managers?
Demand is spread wider than the headlines suggest. The top ten companies account for 153 of the 389 postings, about 39%. The other 76 companies each posted between one and ten PM roles.
Three groups stand out:
- Data and infrastructure platforms. Databricks (20), Datadog (18), Snowflake (13), Nebius (13) and CoreWeave (10) are hiring PMs to put AI into products engineers already pay for.
- Frontier labs. Anthropic (19) and OpenAI (12) are hiring across safety, growth, enterprise and science products. Mistral posted 4 and Cohere 3.
- Agent companies. Sierra (18), Glean (14), Decagon (10) and Harvey (7) sell agents into customer support, enterprise search and legal work. This group is the clearest sign of where the AI PM job is heading.
Scale AI (15) is its own category: most of its PM titles are "AI Product Manager" roles for specific domains, such as Senior AI Product Manager, Finance Agents.
How AllthingsPM does this. Each company in this list that we track has a page of real interview questions, for example Anthropic PM interview questions, and each live JD, such as Product Manager, Growth at Anthropic, opens a mock built from its text. Our decoding posts go deeper on single companies: Anthropic PM job descriptions explained and Scale AI PM roles, decoded.
How senior are AI product manager jobs?
Very. We sorted all 389 titles by keyword:
| Title level (by keyword in the title) | Postings | Share |
|---|---|---|
| Product manager, no level stated | 136 | 35% |
| Senior | 123 | 32% |
| Staff | 74 | 19% |
| Lead, group, head, director or VP | 41 | 11% |
| Principal | 12 | 3% |
| Associate or APM | 3 | 1% |
Source: AllthingsPM JD corpus, 22 September 2026. Titles that mention two levels are counted at the higher one.
Our tagged pass over the 335 AI-native postings, which reads the body and not just the title, came out the same way: 119 senior, 66 staff, 10 principal and only 4 APM. Axial Search's much larger US sample found 2% of AI product postings entry level [1].
The practical reading: the typical buyer is an experienced PM adding AI depth. If you are early in your career, the path in is usually a PM role on a team that ships AI features, not an "AI PM" title on day one. Our guide on how to become an AI product manager walks through that route.
How AllthingsPM does this. The AllthingsPM AI PM course is written for that buyer. It opens with the AI PM job now, a free lesson on why employers want a senior PM who can select, judge and verify AI output, and moves straight to shipping and measuring real systems.
What do AI product manager jobs pay?
71 postings in our archive carried a posted pay range in the job text. These are base salary bands as written by the employer, mostly for US locations. They exclude equity, which at these companies is often a large part of total pay.
| Company | Posted base pay bands seen (USD) | Postings with pay |
|---|---|---|
| Anthropic | $305,000 to $460,000 | 21 |
| Figma | $169,000 to $348,000 | 5 |
| Scale AI | $171,200 to $345,000 | 11 |
| Suno | $200,000 to $330,000 | 4 |
| Harvey | $177,000 to $300,000 | 10 |
| Decagon | $200,000 to $290,000 | 7 |
| Glean | $160,000 to $280,000 | 11 |
| Vercel | $172,000 to $258,000 | 1 |
Source: salary text on each company's own job board, captured by AllthingsPM in 2026. Ranges are the lowest and highest figures across that company's postings, not a single role.
Two patterns are worth knowing before you negotiate. First, the spread inside one company is wide: Scale AI bands start at $171,200 and top out at $345,000 depending on level. Second, the frontier labs pay at the top: every Anthropic PM posting we captured started at $305,000 or more. For more on pay by country, see AI product manager salary in 2026 and AI PM salary in India.
How AllthingsPM does this. Pay follows level, and level is decided in the interview loop. The AllthingsPM question bank holds 4,122 real questions from 260 companies, so you can practice the exact product sense, metrics and strategy questions a senior loop asks, each with an answer guide.
What skills do AI product manager job descriptions ask for?
We tagged each of the 335 AI-native postings against 14 skill themes. A theme counts when the posting asks for it in the responsibilities or requirements.
| Skill theme | Share of JDs | JDs | Companies |
|---|---|---|---|
| AI PM craft (PRD, roadmap, strategy) | 89% | 299 | 83 |
| Technical fluency and ML tradeoffs | 88% | 295 | 85 |
| Enterprise and deployment | 79% | 263 | 77 |
| AI UX and human oversight | 78% | 260 | 80 |
| Agents and agentic architecture | 74% | 248 | 75 |
| 0-to-1 and building under ambiguity | 69% | 230 | 75 |
| Outcomes and metrics for AI | 64% | 213 | 69 |
| Evals and measurement | 53% | 179 | 68 |
| Multimodal (voice, vision, media) | 34% | 113 | 52 |
| Safety, trust and governance | 31% | 105 | 42 |
| Model economics (cost, latency) | 22% | 75 | 42 |
| Context engineering and retrieval | 14% | 46 | 27 |
| Prompting | 10% | 34 | 23 |
| Fine-tuning and customization | 6% | 20 | 13 |
Source: AllthingsPM demand profile, 335 AI-native PM postings at 88 companies, September 2026.
The headline is not that AI PMs need technical fluency; that has been true for years. It is the pair of enterprise deployment (79%) and agents (74%). The median posting wants someone who can ship an agent into another company's systems, with the security review, permissions and admin controls that implies. The top named product surfaces confirm it: "enterprise AI agents" appeared 12 times and "customer support AI agents" 5 times.
Prompting, on the other hand, shows up in only 10% of postings. Employers now assume it.
How AllthingsPM does this. The AllthingsPM course is built from this same demand data, which is why it spends a full chapter on enterprise deployment, the theme most other courses skip. Our analysis of what AI PM courses teach versus what AI PM jobs want shows the gap, and the AI PM skills required in 2026 post turns it into a checklist.
Which tools show up in AI PM job postings?
| Tool named in the posting | Mentions |
|---|---|
| SQL | 28 |
| APIs | 27 |
| MCP (Model Context Protocol) | 25 |
| LLMs | 21 |
| TypeScript | 21 |
| Claude Code | 16 |
| Python | 15 |
| Cursor | 14 |
| Figma, Zendesk, GitHub, Microsoft Teams | 13 each |
| ServiceNow | 12 |
Source: AllthingsPM demand profile, September 2026.
Two shifts to watch. MCP, a protocol for connecting models to tools and data, is now named more often than Python. And AI coding tools are showing up in PM requirements: Claude Code (16), Cursor (14) and Codex (6). Employers increasingly expect a PM to prototype with these tools rather than hand every idea to engineering. We cover what that means in do AI PMs need coding skills and in the full AI PM job postings top tools breakdown.
How AllthingsPM does this. The AllthingsPM course teaches these tools as product decisions, not syntax: when an agent needs a tool connection, what a tool contract should say, and how to prototype a slice before asking for headcount. The knowledge graph shows how concepts like agents, evals and retrieval connect.
Where are AI product manager jobs located?
San Francisco is named in 141 of 389 postings (36%), New York in 90 (23%) and London in 20. Only 67 (17%) mention remote at all, and many of those restrict it to a country or time zone. Many postings list several cities, so these shares overlap.
Axial Search's US sample points the same way: 25% of AI product roles in California and 16% in New York, with 48% hybrid, 31% remote and 20% in person [1]. Lenny's TrueUp data put the San Francisco Bay Area first for PM openings overall, ahead of remote US and New York City [2].
If you are outside these hubs, the honest advice is to target remote-eligible postings first and to treat relocation as part of the offer conversation. Indeed Hiring Lab reports that AI is spreading across job titles well beyond tech occupations [3], so AI PM work at non-tech employers near you is a real second path.
How AllthingsPM does this. Resume Job Match searches open roles by location and ranks them against your resume, and resume review against a JD shows which requirements your resume does not yet prove before you apply.
What changed this quarter, and what to watch next quarter?
This is the first edition, so there is no prior quarter in our own data to compare against. Next edition will report the change on every row of the summary table. Here is what we will watch:
- Junior access. Does the APM share stay near 1%, or do labs open more early-career roles?
- Agents versus chat. Does the agents theme pass 74%, and do "customer support AI agents" keep growing as a surface?
- Builder tools. Do Claude Code, Cursor and Codex mentions keep rising in PM requirements?
- Evals. Evals appear in 53% of postings today. We expect that share to rise as more agents reach production.
- Pay. Do the posted base bands move at the top companies?
The external signals are supportive. Lenny's March 2026 analysis found PM openings at their highest level in over three years [2], and Indeed Hiring Lab found that US job titles with five or more postings mentioning AI grew from 264 in Q1 2022 to 822 in Q1 2026 [3].
How AllthingsPM does this. The AllthingsPM course pipeline re-reads the job corpus and updates lessons weekly, so when a theme rises in this report, the course follows. You do not need to wait for the next edition to prepare for it.
How should you use this report to get an AI PM job?
Turn each section into one action:
- Pick three target companies from the hiring list and read their live postings in full.
- Match your level honestly. If most openings are senior, lead your resume with shipped, measured outcomes.
- Close the top skill gaps: enterprise deployment, agents and evals, in that order of demand.
- Prove one builder tool. A small prototype built with an AI coding tool is a strong portfolio item; see AI PM projects for your portfolio and real examples in PM portfolios.
- Rehearse against the real JD, not generic questions. Run a JD-based mock interview for each target role.
How AllthingsPM does this. Every step above lives in one AllthingsPM account: the catalog for step 1, resume review for step 2, the course for step 3, portfolios for step 4 and JD mocks for step 5.
Why AllthingsPM is the better choice for AI product manager job prep
A hiring report is only useful if you can act on it the same day. AllthingsPM is built on the same job data this report uses, so the path from "who is hiring" to "am I ready for that loop" is one click.
The advantages are concrete. AllthingsPM keeps 116 live PM job descriptions at 18 AI companies, and each opens a scored mock interview built from that JD, in text or voice, with follow-ups. The question bank has 4,122 real questions from 260 companies, each with its own page and answer guide. The AI PM course was built from 604 real PM job postings and has 14 chapters, 101 lessons and 14 graded case studies, updated weekly. Resume review against a JD, Resume Job Match, 111 book summaries, podcast summaries and 455 PM portfolios sit in the same account.
Job boards such as LinkedIn and aggregators have far more listings; for AI PM roles specifically, AllthingsPM gives you the posting, the mock built from it and the course that closes the gap, together. Cohort courses offer live instruction; for daily preparation at $20 a month or $120 a year, with a free JD mock every day, AllthingsPM is the practical choice.
The verdict: use this report to choose targets, then start free on AllthingsPM to prepare for them.
Frequently asked questions
What is the best way to find AI product manager jobs?
The best place to start is AllthingsPM, which lists 116 live PM roles at 18 AI companies and builds a mock interview from each job description. Then read company job boards directly, since that is where most AI companies post first, and use Resume Job Match to rank open roles against your resume.
How many AI product manager jobs are open right now?
On 22 September 2026, AllthingsPM found 389 open PM postings at 86 AI companies across 160 job boards. Wider counts are larger: Axial Search tracked 12,397 US AI product postings in 2026, and Lenny Rachitsky counted over 7,300 open PM roles at tech companies globally in March 2026.
Are there entry level AI product manager jobs?
Very few. Only 3 of the 389 postings we read were associate or APM titles, and Axial Search found 2% of US AI product postings entry level. Most people move into AI PM work from a PM role on a team that ships AI features.
How much do AI product managers make?
Posted base pay bands at AI companies ran from $160,000 to $460,000 in our data, before equity. Anthropic bands started at $305,000; Glean's started at $160,000.
What skills do AI product manager jobs require most?
PM craft (89% of postings), technical fluency (88%), enterprise deployment (79%), AI UX (78%) and agents (74%). Evals appear in 53%. Prompting appears in only 10%, because employers now assume it.
How often is this report updated?
Quarterly, four times a year, with the same sections in the same order so each quarter can be compared with the last.
Start preparing today
Pick one role from this report, open it in the AllthingsPM jobs catalog and run your free JD mock interview today.
Sources
- Axial Search, "AI Product Management Jobs in 2026: What 12,400 Postings Reveal": https://axialsearch.com/insights/ai-product-jobs
- Lenny Rachitsky, "State of the product job market in early 2026", Lenny's Newsletter, 24 March 2026: https://www.lennysnewsletter.com/p/state-of-the-product-job-market-in-ee9
- Indeed Hiring Lab, "AI Is No Longer Just a Tech Occupation Story", 8 July 2026: https://hiringlab.indeed.com/2026/07/08/ai-is-no-longer-just-a-tech-occupation-story/
- Anthropic job board (Greenhouse), PM postings and pay ranges: https://job-boards.greenhouse.io/anthropic
- Scale AI job board (Greenhouse): https://job-boards.greenhouse.io/scaleai
- Glean job board (Greenhouse): https://job-boards.greenhouse.io/gleanwork
- Harvey job board (Ashby): https://jobs.ashbyhq.com/harvey
- Decagon job board (Ashby): https://jobs.ashbyhq.com/decagon
- Sierra job board (Ashby): https://jobs.ashbyhq.com/sierra
- AllthingsPM JD corpus and demand profile (389 postings at 86 companies from 160 job boards, read 22 September 2026), published as the AllthingsPM jobs catalog: https://allthingspm.app/jobs
