From Core PM to AI/ML PM: Two Ways AI Is Changing Product Management

"AI" already appears in 21% of core PM job postings across the openings we track. One in five core PM roles is explicitly referencing AI as a required or preferred skill, and that share is climbing. Technical PM roles sit even higher, with 28% mentioning AI and 10% specifically calling out LLMs.

This creates an interesting split in what "AI product management" actually means. There are two distinct dimensions emerging, and the data shows them clearly.

Two dimensions of AI in product management

The first dimension is AI as a PM tool. Every product manager is increasingly expected to use AI in their workflow: generating user research summaries, prototyping with LLMs, using copilots for data analysis, accelerating spec writing. This is the 21% signal in core PM roles. It's becoming table stakes for the profession, regardless of what product you're building.

The second dimension is AI as the product. This is what AI/ML PM roles are really hiring for: someone who can make product decisions about machine learning systems. Should we fine-tune or use RAG? How do we evaluate hallucination rates? What's the latency budget for inference? When do we retrain? These are questions that require a deeper understanding of how AI systems work under the hood.

The skill data draws this distinction sharply. Nine of the fifteen most in-demand skills for AI/ML PM roles are standard core PM skills: SQL, data analysis, stakeholder management, product strategy, roadmapping, user research, metrics, cross-functional collaboration, and AI itself. The product management fundamentals carry over completely.

The six bridging skills all sit in that second dimension: LLMs (28% of AI/ML PM roles versus 2% of core PM), machine learning (18% versus 4%), generative AI (15% versus 2%), prompt engineering (10% versus under 1%), AI agents (9% versus under 1%), and Python (13% versus 8%). These are the skills you need to make informed product decisions about AI systems, to understand what the engineering team is building well enough to shape it.

Python at 13% reinforces something important about the second dimension. It's useful for prototyping, querying data, and working alongside ML engineers in notebooks. But AI/ML PM roles are asking for conversational fluency with the technical stack rather than deep engineering ownership. Compare Python at 13% for AI/ML PM with 63% for ML engineers. The expectation is that you can evaluate the technology, not build it.

What shifts in emphasis

When job descriptions allocate more space to AI domain knowledge, something has to give. The data shows what: Agile appears in 25% of core PM roles but only 11% of AI/ML PM roles. JIRA goes from 16% to below the top twenty. Scrum drops from 9% to negligible.

AI/ML PMs still run sprints and use project management tools. But when companies write these job descriptions, they emphasise the ability to evaluate an LLM's output quality, define success metrics for a recommendation engine, or make build-versus-buy decisions on model infrastructure. The process skills are assumed. The overlap skills themselves tell the same story: most appear at lower frequency in AI/ML PM postings (SQL drops from 28% to 22%, data analysis from 26% to 16%), likely because job descriptions spend their word budget on the AI-specific requirements they can't take for granted.

Who companies are hiring

AI/ML PM roles skew notably toward the senior end. Staff and principal level roles account for 21% of AI/ML PM openings, more than double the 10% for core PM.

This tells you who companies want for these roles: experienced product people who can layer AI expertise onto a strong PM foundation. If you're a mid-level or senior core PM, the market values your product experience as a prerequisite, then looks for AI literacy on top of it.

The industry mix spreads more evenly than core PM. Where core PM roles concentrate in fintech (24%), AI/ML PM hiring distributes across AI-native companies (13%), fintech (13%), productivity tools (12%), and data infrastructure (10%). San Francisco accounts for 39% of AI/ML PM openings, more than double its 15% share of core PM roles, though New York (26%) and London (14%) both have meaningful volume.

In the US market, AI/ML PM roles range from around $190k to $265k. Core PM roles sit at $168k to $229k. The premium is consistent across the salary band, roughly $20k to $35k higher at each quartile.

Practical next steps

The two dimensions suggest a natural sequencing.

Start with the first dimension: become the PM who uses AI fluently. Use LLMs daily for real work. Build prompts that solve actual problems in your current product. Learn what makes outputs better or worse. This investment pays off regardless of whether you pursue AI/ML PM roles, because AI literacy is becoming expected across all PM positions.

Then move to the second dimension: develop the domain knowledge to manage AI products. The biggest skill gaps are conceptual: LLMs, machine learning, generative AI. You need to understand how these systems work, what they're good at, where they hallucinate or fail, and how to evaluate their output. This requires spending serious time understanding the product trade-offs these technologies create (latency vs accuracy, cost vs quality, deterministic vs probabilistic). Prompt engineering, which appears in 10% of AI/ML PM roles, is one of the most immediately applicable bridges between the two dimensions.

Ship something with AI. The gap between "I understand LLMs conceptually" and "I've built a product feature powered by an LLM" is where interviews separate candidates. Build a side project, add an AI feature to an existing product, or volunteer for AI initiatives at your current company. The portfolio signal matters more than the certificate.

You already have the product craft: strategy, stakeholder management, roadmapping, user research, data analysis. These are the foundation that companies hiring AI/ML PMs explicitly value. The AI-specific skills are learnable, the market is paying a clear premium for them, and the direction of travel suggests that AI literacy will only become more important across all product roles, whether or not your title includes "AI."

---

Methodology: This analysis draws on approximately 4,300 product management job postings collected between November 2025 and February 2026 from company career pages. Postings span London, New York, San Francisco, Denver, and Singapore. Skills were extracted from full job descriptions using an LLM classifier. Salary data is limited to US markets where disclosure is more common. Recruitment agency listings and out-of-scope roles were excluded. For skills analysis, the core PM sample is 718 postings and the AI/ML PM sample is 231 postings. Salary sample sizes: core PM n=256, AI/ML PM n=93.