Product managers are some of the fastest AI adopters in tech. But according to some of our new research, even as nearly all PMs rely on AI daily, most are learning as they go, with a surprising number going off the books to do it.
Our recent survey of 117 product managers across the US, UK, Canada, and Singapore reveals a clear truth: AI has become an everyday tool of the trade. But the skills, policies, and training behind it still need to catch up.
AI is embedded in product management workflows
Nearly every product manager surveyed (98%) said they use AI at work. On average, they tap into AI tools 11 times a day, while those in the top tenth percentile use it as often as 25 times daily.
And this isn’t just casual use. 78% of product managers already use AI agents to streamline tasks, automate steps in product development, or accelerate decision-making.
Despite how widespread AI has become in product management and beyond, governance remains murky. While 62% say their companies provide sanctioned AI tools, two-thirds (66%) admit to using unapproved options—a growing trend known as shadow AI.
When it comes to AI for product management: the enthusiasm is there, but the guardrails aren’t.
Most PMs are encouraged to use AI—but few are required
AI is no longer fringe in product teams. In fact, 65% of PMs say they’re encouraged to use it, and only 3% say they’re discouraged from doing so.
But formal mandates remain rare, with just 9% saying their company requires AI use. The rest are encouraged, allowed, or experimenting independently.
When asked why they use AI, productivity and curiosity top the list:
- 23% want to be more productive
- 18% want to reduce workload
- 18% simply enjoy experimenting with new tools
- 14% want to improve their work quality
- 13% are motivated by skill development
That curiosity is shaping the next phase of product management—where PMs aren’t just managing roadmaps, but also the intelligent systems that shape them.
Role-specific AI training is the missing piece
The most striking stat from the survey may be this: only 39% of PMs say they’ve received comprehensive, job-specific AI training.
Another 19% say their training was too generic, and another 19% say it was basic and conceptual. Add it up and this means that less than half of product managers feel well prepared to use AI effectively at work.
Instead, many are taking matters into their own hands:
- 45% have learned AI independently
- 29% attended company-funded external programs
- 20% paid out of pocket for training
And just 3% say they’ve received no AI training at all.
As our Director of Product Management Beatrice Partain puts it:
The complexity, scale, and accountability of PM work exceed what informal learning can reliably provide. Job-specific training can turn experimentation into scale—taking AI skills beyond surface-level productivity boosts to accelerate strategic decisions and innovation.
This highlights a clear inflection point for companies: AI adoption has outpaced enablement.
PMs want training that evolves with AI
When asked what kind of AI learning would be most valuable, product managers were clear that one-and-done training doesn’t cut it.
They want:
- 64% – Regular training updates as AI tools evolve
- 51% – Peer learning sessions with colleagues
- 49% – Self-paced modules with PM-specific examples
- 40% – Ongoing support and troubleshooting
- 37% – Interactive workshops on specific use cases
The takeaway: PMs don’t just want to “check the box” on AI training. They want continuous learning that matches the pace of change.
AI use cases are expanding quickly
Product managers aren’t just using AI for brainstorming or copywriting, they’re using it across the entire product lifecycle.
The top use cases include:
- Managing product development cycles, sprint planning, and delivery (54%)
- Cross-functional collaboration (52%)
- Creating product strategies and roadmaps (48%)
- Conducting or simulating customer interviews (46%)
- Backlog grooming and QA support (44%)
- Analyzing customer feedback (42%)
They’re even starting to build and deploy advanced tools, with 31% designing custom AI agents or domain-adapted models tailored to their products.
But there’s the catch: interest exceeds expertise. For nearly every task, more PMs want to learn how to use AI for it than currently know how.
For example:
- 61% want to use AI for managing product development cycles, but only 54% do
- 47% want to prototype and validate products without engineering (aka they want to vibe code), but just 38% can
That gap represents an opportunity for companies: supporting structured AI upskilling could unlock the full productivity gains PMs are already chasing.
AI is boosting productivity, not cutting headcount
Despite fears about automation replacing jobs, the data tells a different story for now.
- 66% of PMs say their teams are the same size but more productive thanks to AI
- 26% report their teams have actually grown since adopting AI
- Only 1% say there are fewer people on their teams since AI was introduced
The AI shift isn’t about replacement—it’s about leverage. Teams are learning to use AI to accelerate delivery, scale insights, and focus human energy on more strategic work.
AI is changing how PMs work and think
Nearly every product manager in the survey agrees that AI is improving their workflows:
- 98% say it’s improved the product lifecycle
- 97% say it helps their department make faster decisions
- 75% say it’s freed them to focus on more strategic priorities
- 40% say they’re working fewer hours because of it
Still, not all impacts are positive. 15% say AI has added to their workload or pulled them away from strategic projects. And even with all the upside, there’s a layer of unease underneath.
When asked about their biggest concerns:
- 26% fear AI could eventually replace them
- 25% worry it will make it harder for entry-level PMs to learn
- 22% fear it could replace colleagues
- 15% say it could make work less enjoyable
So while AI may not be taking PM jobs, it’s reshaping what the role looks like, and what it takes to thrive in it.
The future: from “experimenters” to AI strategists
Our latest survey paints a picture of product managers who are eager, capable, and curious, but also at a crossroads. They’re among the earliest adopters of AI at work, yet they’re doing it in an environment where skills and policies are still evolving in real time.
In fact, 100% of PMs say their leaders expect them to use AI before asking for more resources. That mandate puts PMs on the front lines of AI adoption and makes structured upskilling, like our AI for Product Managers Workshop, a strategic necessity, not a nice-to-have.
AI isn’t just changing what product managers use to get work done. It’s changing what they need to know to lead.
And as our research shows, the next leap forward for product teams won’t come from more AI tools. It’ll come from building the skills to use them well.
About the Survey
General Assembly surveyed 117 product managers in the United States, United Kingdom, Canada, and Singapore between October 2–13, 2025. Respondents worked at companies with 100+ employees, managing software and digital service products.