According to Maze’s Future of User Research Report 2025, 58% of product teams now use AI tools for research. That’s a 32% increase year-over-year. The adoption curve isn’t slowing — it’s steepening.
On the surface, this looks like progress. More teams using AI means more efficiency, more insights, faster decisions. Right?
Not necessarily.
The Two Types of AI Research
There’s a critical distinction that most teams aren’t making. AI in research falls into two very different categories:
The first is AI as accelerator. This is where 74% of adopters are using it — analyzing existing research data, transcribing interviews, synthesizing notes. AI takes real human insights and processes them faster. This is genuinely valuable.
The second is AI as generator. This is where teams use AI to create research output directly — synthetic personas, simulated interviews, AI-only “validation.” No real users involved.
The difference matters enormously. One starts with human truth and uses AI to make it accessible. The other starts with AI assumptions and dresses them up as human truth.
The Speed Trap
Here’s what drives the shift from accelerator to generator: time pressure.
63% of product teams cite time and bandwidth as their top challenge (Source: Maze 2025). When you’re moving fast and your research team is outnumbered 1:50 by developers (Source: NNg), the temptation to let AI “handle” the research is enormous.
And AI makes that temptation easy to give in to. It produces professional-looking output in minutes. No recruiting. No scheduling. No messy human complexity.
But professional-looking isn’t the same as valid. And fast isn’t the same as right.
The 58% Question
The real question isn’t whether your team is using AI for research. By now, most teams are. The question is how.
Are you using AI to get closer to your users — or as a substitute for them?
Are you accelerating real insights — or generating comfortable fictions?
The answer determines whether you end up in the 5% of products that succeed or the 95% that don’t (Source: Clayton Christensen, HBS).
Building the Right Tool
This distinction is why we built Seena the way we did. Not AI that replaces research. AI that makes real research faster, more connected, and impossible to ignore.
Real user signals, amplified by intelligence. Not synthetic signals dressed up as truth.
The teams that win the next decade won’t be the ones who adopted AI first. They’ll be the ones who adopted it correctly.
Join the waitlist at SeenaLabs.io
—Ax



