Part 1 of 3. A JobMarketPulse case study of how four major tech companies rewrote existing roles for AI.
We read 2,409 job postings for Product Manager, Analyst, and Marketing Manager roles at Apple, Google, Microsoft, and Nvidia, published between 2023 and 2026. We pulled them from JobMarketPulse, our job postings platform, and counted word frequency, identified new job titles, and noted required degrees.
This is a case study of four companies, not a picture of the whole job market. All four build AI products, which makes their postings a useful early look at how AI changes what a job description asks for.
The short version: Apple, Google, Microsoft, and Nvidia mostly did not create new AI jobs. They added AI requirements to the jobs they already had. The AI word that spread fastest across their postings wasn’t a product name. It was “agentic.”
How we built this
Every posting comes from JobMarketPulse, which tracks the career sites of more than 400,000 companies worldwide. For this series, we picked four of them and read every Product Manager, Analyst, and Marketing Manager posting they published from 2023 to 2026.
A figure like “19.2% of postings” means 19.2% of that role’s postings from all four companies in that year, unless we name a single company. The year-by-year numbers come from charts we built and checked. The title counts by company come from summary totals, so we’ve checked those less closely.
“Agentic” went from almost nothing to common in two years
“Agentic” describes AI that carries out several steps on its own to reach a goal, instead of answering one prompt at a time. “Agent” is the plainer word for the same idea. In 2023 and 2024, almost no postings at the four companies used either one.
By 2026, “agentic” appeared in 19.2% of Product Manager postings, 6.8% of Marketing Manager postings, and 4.4% of Analyst postings. “Agent” went further. It rose from 0% in 2023 to 29.0% of Product Manager postings, 10.1% of Analyst postings, and 9.7% of Marketing Manager postings.
Figure 1: Agentic and agent mentions surge across Product Manager, Analyst, and Marketing Manager postings, 2023-2026
The chart combines all four companies, so it can’t show who moved first. We can split it by company only for Microsoft and Nvidia, and they moved in different years. At Microsoft, “agent” was already in 25.9% of Product Manager postings in 2025. At Nvidia, it stayed at 0% until 2026.
A caution: the word may be ahead of the work
Gartner, a research and advisory firm, has warned about “agent washing,” where vendors relabel existing products as “agentic” without changing much underneath. It predicts that more than 40% of agentic AI projects will be canceled by the end of 2027 because of cost, unclear business value, and weak risk controls.
The same thing can happen in job postings. A word can spread faster than the work behind it. Our data shows the word is common at these four companies. It can’t show that every one of those jobs involves something an agent-powered system couldn’t have done in 2023.
New AI job titles stayed rare
At all four companies, titles like AI Product Manager, AI Analyst, and AI Marketing Manager made up under 5% of their role postings in every year from 2023 to 2026. “AI/ML Product Manager” appeared in 36 of 1,205 Product Manager postings, just under 3%, mostly at Google and Apple. “AI Marketing Manager” appeared in none of the 274 Marketing Manager postings. The AI requirements showed up inside the regular Product Manager, Analyst, and Product Marketing Manager postings instead.
Two exceptions are worth naming. Microsoft introduced 20 Product Manager titles branded with Copilot, its AI assistant, in 2025. That is product branding more than a new kind of job. And Technical Product Manager postings grew from 4 in 2023 to 14 in 2026, mostly at Nvidia and Google. That may be a real shift, but 4 to 14 is a small base, so a few postings either way would change the picture.
Nvidia and Microsoft show the most AI language of the four in our data, and even they mostly rewrote existing roles instead of creating new ones.
Figure 2: Product Manager title subtypes by year, showing AI/ML PM and Technical PM staying thin next to generic, senior, and principal PM
A company’s own product name peaks, then the general word keeps going
At Microsoft, “copilot” appeared in 66.7% of Product Manager postings in 2025, then dropped to 39.2% in 2026. “Agent” kept climbing, from 25.9% to 52.9%.
Figure 3: Microsoft Product Manager postings: AI, Copilot, agent, and agentic mentions, showing the Copilot spike and pullback
Nvidia shows the same shape with “GPU,” the chip it sells. In Product Manager postings, “GPU” appeared in 93.3% of postings in 2025 and 66.7% in 2026, while “agent” went from 0% to 48.1%. In Marketing Manager postings, “GPU” went from 81.8% in 2025 to 62.5% in 2026, while “AI” stayed at 100% in both years.
Figure 4: Nvidia Marketing Manager postings: AI, GPU, and agent mentions, showing the GPU spike and pullback
We call this the brand-overshoot pattern: a company’s own product name peaks in its job postings, then slips while a more general AI word holds or grows. We see it at Microsoft and Nvidia. We’d want to see it at other companies before calling it a general pattern.
We don’t have the same product-name view for Google and Apple. Their general AI mentions in Product Manager postings rose gradually between 2023 and 2026, from roughly 5% to 50% at Google and roughly 6% to 46% at Apple.
Frequently asked questions
What does “agentic AI” mean in a job description? It means the posting expects the person to build, manage, or work alongside AI that takes multi-step actions on its own, not just answers a prompt. At Apple, Google, Microsoft, and Nvidia, the word barely appeared before 2025 and was in 19.2% of Product Manager postings by 2026.
Are Apple, Google, Microsoft, and Nvidia creating new job titles for AI? Mostly no. AI-branded titles stayed under 5% of each role’s postings in every year from 2023 to 2026. The AI requirements went into the existing Product Manager, Analyst, and Marketing Manager titles.
Which of the four companies mentions AI the most in job postings? Nvidia, then Microsoft. In 2026, “AI” appeared in 100% of Nvidia’s Product Manager postings and 90.2% of Microsoft’s, compared with roughly 50% at Google and 46% at Apple.
What comes next in this series
This post covered what changed across all three roles. In Part 2, we look at the Product Manager job description in detail: which skills dropped out, which moved in, and how degree requirements shifted. In Part 3, we cover Analyst and Marketing Manager, where the story gets messier. Read Part 2 | Read Part 3
Want this for your own competitive set?
JobMarketPulse tracks more than 400,000 companies and 11.5 million live jobs worldwide, including 6.4 million live jobs in the U.S., about 70% of the U.S. market. Our JobsIndex database goes back to September 2022 and refreshes daily. If you want this kind of analysis run against a different set of employers or roles, reach out to our research team, and we’ll scope it out.
About Aspen Technology Labs
Aspen Technology Labs, Inc. is a global leader in recruitment technology, labor market and wage analytics, and web data management services. Founded in 2008 and headquartered in Aspen, Colorado, Aspen’s JobMarketPulse platform delivers real-time labor market intelligence built to be easy to use and understand, no data science degree required. As of October 6, 2026, Aspen collects job posting data from more than 400,000 companies across 195 countries, including more than 260,000 with jobs posted in the U.S., and serves more than 300 customers.
See this data in action
Explore structured job data and labor market intelligence for your use case.
