-

AnswerLab Research Finds People Can Describe What AI Is Costing Them, as They Keep Using It Anyway

Based on 334 interviews across seven areas of life, the research finds people are writing their own rules for AI and can describe exactly what it’s costing them while they keep using it

NEW YORK--(BUSINESS WIRE)--AnswerLab, the research-powered experience strategy firm, today released its signature report, The Assumption Gap: How People Are Setting Their Own Rules for AI. The report finds a growing gap between what AI products assume people want and what people actually do, with some of the most notable evidence coming from those who can describe how AI is wearing down their effort and confidence as they keep using it anyway.

AnswerLab drew on three years of insight from hundreds of research engagements, a review of public research, and conversations with experts in patient advocacy, vehicle safety, machine learning, and digital trust. It then ran seven studies, one each in Commerce, Health, Home, Relationships, School, Transportation, and Work. Via AI-moderated interviews, 334 people – including two cohorts of teens – described how they use AI, where they stop, and what they do when products don’t support the distinctions they need.

Across every domain, from grocery shopping to healthcare to the classroom, participants wanted AI to research, compare, and draft, but not decide. For instance, identical language across categories included: "I want to approve it myself, " and "That's where I draw the line." The most experienced users were most likely to build their own rules, down to exact dollar thresholds and category exclusions.

"People are already setting the rules for how AI fits into their lives, purchase by purchase, message by message, action by action, without any company asking them to," said Megan Malli, CEO at AnswerLab. "People aren’t waiting for anyone to hand it to them and, frankly, are showing us that they’ll use AI, but on their own terms, in a way they’re comfortable with after building that sense of trust. The companies that build for that instinct, instead of trying to design around it, are the ones that will earn trust fastest."

The Confidence Trap: Teens Can Name What AI Is Costing Them

Heavy AI users – teens in particular – can describe exactly how AI is changing their thinking, and they still can’t stop using it. One teen participant put it plainly: “I’m losing every sense of creativity. I’m also losing the habit of effort.”

But the cost goes beyond lost skills. One student called the tradeoff a wash, then minutes later said: “I feel [kind] of scared without AI, especially if I have to do a coding assignment and I don’t have AI to help me. I just kind of dread it.”

The difference came down to how people used the tool. Across the study, participants who used AI to understand a topic, test results, a lease, or a homework concept described feeling more prepared and more willing to push back. Those who used it to skip the work described dependency. The same technology built confidence when it explained and wore it down when it did the work.

Managing AI Has Become a Second Job

Introducing AI into workflows was intended to automate tasks, but across the study, people described managing AI as an unaccounted-for second job: briefing AI, coordinating it across tools, checking its output, and keeping it running.

Roughly 80% of participants across five study domains described a habit of checking AI's work. In Health and School, where the question was asked directly, the number was highest. In Work, Commerce, and Home, participants volunteered the behavior unprompted.

In the Home study, coordination was the most widely shared complaint. One participant counted roughly ten apps running their house. Another builds routines across devices by hand because "none of those apps are really trying to make things easier for me." The promise is a home that runs itself in the background, but the reality is a stack of separate tools connected by the person who lives there.

It would be easy to read all this as a learning curve, the kind of effort that fades as people get comfortable with a new tool, but the evidence suggests otherwise. The prompts people write have gotten longer over time, not shorter, and at work, time saved turned into higher expectations, so the management never ramped down.

The Effort Was the Point

When people spot AI in a personal message, what bothers them is the missing effort.

About half of the teens in the Relationships study said they noticed AI help in messages and didn’t mind, but that tolerance has a clear limit. Teens welcomed AI in formal messages and protected the personal ones. Some refused to respond to a personal message they thought was AI-written. One said an AI-written paragraph from an ex trying to reconcile “counted for nothing.”

People caught AI-written messages less by how polished they were than by how generic they sounded and how little they sounded like the sender. Many who use AI to draft their own messages described editing them back into their own voice. As one put it: "Those who know me know all my grammar and punctuation isn't that perfect."

The New Front Door

By the time customers arrive at a business, many have already had something close to a consultation with their own AI, and the company's front door may never open.

Roughly 70% of Commerce participants found AI review summaries useful for making purchase decisions faster. Nearly half described using AI to plan travel by compiling itineraries, amenities, and prices across a variety of online sources.

Participants want AI to do more of the homework: hunt prices across retailers, verify sellers, and track loyalty rewards. But ultimately, they want to continue making the decisions themselves.

More than 20% of Commerce participants – unprompted – described customer service chatbots as frustrating, circular, or "a dead end." What bothered them was less what the bots couldn't do than the fact that the bots’ answers served the company’s interests over the consumers’.

For some, the first stop is becoming the only stop. One participant checked her own blind spots with a general-purpose AI instead of calling a financial advisor. Another admitted to using AI in place of the freelancers and interns he used to hire.

These findings point to a shift in how companies should think about AI adoption. The dividing line has progressed beyond AI skeptics and enthusiasts. It's now between companies that pay attention to what their customers are actually using AI for and those that aren’t. The workarounds in this report are a map of where AI products need to go next.

Download the full AnswerLab Signature Report here.

Research Methodology

The AnswerLab Signature Report was conducted between July 1 and July 13, 2026. The study included 334 qualitative interviews across seven life domains: Commerce, Health, Work, School, Home, Relationships, and Transportation, including two under-18 cohorts, across the US. AI moderation made it possible to conduct this research at scale while still asking the follow-up questions that traditional surveys never get to. The result is a dataset that is both wider than a typical qualitative study and deeper than any check-the-box survey could produce. AnswerLab paired the interviews with three years of client research and interviews with outside experts in patient advocacy, vehicle safety, machine learning, and digital trust.

About AnswerLab

AnswerLab is a research-powered experience strategy firm that helps pioneering leaders make confident decisions in moments of uncertainty. By combining deep human insight, AI-enabled research, rapid prototyping, and validated roadmapping, AnswerLab enables organizations to reduce risk, sharpen strategy, and design experiences that drive meaningful business impact. Trusted by many of the world's most influential brands, AnswerLab partners with clients globally across technology, financial services, healthcare, retail, and emerging sectors.

FAQ:

What is the assumption gap in AI?
The assumption gap is the distance between what AI products assume people want and what people actually do. AnswerLab’s research finds the clearest evidence of it in the workarounds people build to close the gap themselves.

What did AnswerLab’s research find about teens and AI?
Teens could describe precisely how AI was affecting their effort, creativity, and confidence, and many kept using it anyway. Some turned AI into a tutor, telling it not to give them the answer and to help them understand the problem instead.

Do people trust AI to make purchases on their own?
Not for anything new. None of the 50 Commerce participants in AnswerLab’s study would let AI make a novel purchase on its own, at any price. About 40% would let AI buy within limits they set.

Do people check AI’s work?
Yes. Roughly 80% of participants across five of AnswerLab’s studies described checking AI’s answers, treating it as they would review an intern’s draft. That checking tends to fade as AI builds a track record, just as a missed error starts to cost more.

How was the AnswerLab Signature Report conducted?
AnswerLab conducted 334 AI-moderated qualitative interviews across seven areas of daily life, including two teen cohorts, and paired them with three years of client research and interviews with outside experts.

Contacts

Media Contact

Unshakable Marketing Group for AnswerLab
answerlab@unshakablemarketinggroup.com

AnswerLab


Release Summary
AnswerLab released its signature report, The Assumption Gap: How People Are Setting Their Own Rules for AI.
Release Versions

Contacts

Media Contact

Unshakable Marketing Group for AnswerLab
answerlab@unshakablemarketinggroup.com

More News From AnswerLab

AnswerLab Bolsters Experience Strategy Offering with Key Leadership Moves

NEW YORK--(BUSINESS WIRE)--AnswerLab, a research-powered experience strategy firm, today announced the appointment of Kevin Kearney as Chief Experience Officer and the promotion of Ashley Craig to Managing Director, Research & Product Innovation. These moves further strengthen the company’s research, experience strategy, and innovation capabilities. Kearney and Craig will report to CEO Megan Malli. As Chief Experience Officer, Kearney will spearhead AnswerLab’s Experience Strategy and Desig...

AnswerLab Unveils a New Chapter—Transforming Experience Strategy to Meet the Demands of What’s Next

NEW YORK--(BUSINESS WIRE)--AnswerLab today announced a significant shift from traditional UX research to a research-powered experience strategy and design firm....
Back to Newsroom