Twice as many consumers grant AI access reluctantly as willingly. Call it resigned consent.

Close to six in ten consumers say they would let an AI assistant access their data: their work tools, their health data, their company account information. On paper, that looks like a win. Consumers are on board and adoption is coming.

New data from Usercentrics’ State of Digital Trust 2026, based on 11,000 consumers across seven markets, says that’s only half right.

An average of 58% of consumers report some level of discomfort with AI accessing their data across six categories, ranging from 53% for work tools to 64% for financial accounts. Only 7% are fully comfortable granting access with no conditions attached, and 16% will tell you they’re granting that access despite feeling uncomfortable about it.

This is called resigned consent. The person said yes because they’d already decided no wasn’t a real option. It outnumbers willing access two to one across every category and every generation surveyed. The only thing that changes that picture is country, a difference we will explore in this report.

A consent rate tells you people said yes. It doesn’t tell you which yes you got. Genuine consent shows up as loyalty and retention. Resigned consent shows up as churn and complaints, without anyone tracing either one back to the original opt-in. Distinguishing between the two happens later, in usage data, renewal conversations, and the permissions people choose to turn off.

Closing the gap means treating an access rate as the start of the work rather than proof it’s done. Access isn’t the finish line, it’s the opening move in a relationship, and what a brand does with that access afterward is what turns a resigned yes into a genuine one.

Not all resignation looks the same. Financial accounts, the highest-stakes category tested, produce the most discomfort among consumers surveyed, at 64%. But resigned consent there is the lowest of any category, at 14%. When the data is sensitive, people don’t grudgingly say yes. They leave: almost a quarter would stop using the AI entirely rather than give it access to their financial accounts.

Work tools flip that completely. Discomfort drops to 53%, walking away goes down to 13%, while resigned consent holds at 17%, level with email and health data. Work tools is where AI adoption is actually happening right now, the copilots, the assistants, the agents companies are rolling out to their own people. It’s also the single category where the least discomfort turns into a decision; people aren’t refusing AI at work, they’re just not meaning the yes.

But that discomfort doesn’t disappear, it just turns into a harder no elsewhere in the user journey. The bigger the ask, the less likely people are to shrug and go along with it.

Resigned consent peaks at 17%, in work tools, email and health data, with company data right behind at 16%. What connects them might not be obvious from this data alone, but noting they’re also the four categories with the lowest total discomfort paints a more complete picture.

It’s the everyday, low-friction categories where people are most likely to swallow their discomfort rather than act on it. That pattern lines up with what design researchers call consent fatigue: the more frequently someone gets asked the same kind of question, the less they engage actively with it. The repetition wears down the will to keep responding, but the concern doesn’t go away.

At the other end, the stakes are high enough to interrupt the habit. A request for access to someone’s financial data gets read carefully instead of clicked through, and a decision is made – usually a no.

But resigned consent is still the smaller piece of the picture. Most of the discomfort with data sharing expressed by respondents doesn’t turn into a reluctant yes at all. Both flat refusal and walking away from the product already show up somewhere: a declined permission, a lost user. Resigned consent doesn’t show up anywhere. It just looks like a yes.

That’s what makes it the one form of discomfort that quietly makes it into your adoption numbers as a win. It isn’t one, it’s a risk that holds only until something breaks it: a data breach, a change to the privacy terms, a feature that quietly expands what it accesses over time.

“A resigned yes and a real one look identical in your consent rate, but behave nothing alike. One holds under pressure. The other holds until the first bad headline. If you can’t tell which one you have, you don’t know how much of your AI adoption is actually yours to keep.”

Tilman Harmeling

— Strategy & Market Intelligence, Usercentrics

At 24% on average, Germany’s rate of resigned consent is more than double the UK’s or the US’s rate. The average is not being dragged up by one particular category, it holds between 23% and 24% in all six categories. Germany is also the market least likely to say no outright, with 16% flat refusal against a 23% global average, the lowest refusal rate in every category tested. 

Germany is the market most ready to say yes, and that yes is the most reluctant of any market in the study.

The UK and the US sit at the other end. Their rates of resigned consent are the lowest in the study, at 11% to 10% against Germany’s 24%, so when consumers in those markets say yes, they’re more likely to mean it. A lower access rate in those markets isn’t a weaker result, it’s a clearer one.

The State of Digital Trust 2026 helps explain the gap in detail: German consumers rank trust in the regulatory framework above brand transparency as their reason for granting access, the opposite of every other market studied. That’s plausibly why comfort there runs high but genuine enthusiasm doesn’t: people are relying on the rules to catch problems, not on the request itself feeling right.

Discomfort with AI access doesn’t fade evenly as people get older. Refusal and abandonment jump sharply after 45. Resigned consent declines more gradually.

Gen Z and Millennials behave almost identically: tied at 17% on resigned consent, one point apart on refusal, and tied again at 14% on abandoning the product.

Past 45, refusal and abandonment both climb hard. Refusal rises to 26% for Gen X and 30% for Boomers+, up from 20-21% under 45. Abandonment nearly doubles after the threshold, from 14% to 18% to 27%. Resigned consent moves the other way, dropping to 13% among the over-60s.

Both age blocs are uncomfortable. They just resolve it differently: under-45s stay and resign, over-45s are more likely to refuse outright or leave.

Here’s the twist: once someone actually says yes, age barely matters. Look only at the yeses each generation gives, and roughly a quarter to a third of those are reluctant, in every generation, Gen Z included. Age doesn’t decide how reluctant a “yes” is. It decides how often someone gets to yes in the first place.

Millennials and Gen Z are the least uneasy generations in the dataset, essentially tied on both measures. Almost a third of Boomers+ (30%) trust AI less than humans with their personal data, and 37% find AI personalization intrusive.

That tracks with behavior in one specific way: the groups reporting the least unease are also the groups least likely to refuse outright. It doesn’t mean their access is less reluctant. Once they say yes, the reluctance behind it looks about the same as everyone else’s.

Two people granted an AI assistant access to their inbox this morning. One weighed the trade and decided it was worth it. The other decided refusing was more trouble than allowing it. Your access rate recorded them as the same.

There’s no single fix for that. How people resolve their discomfort varies by market and by age. German consumers grant access reluctantly. British and American consumers grant it less reluctantly, but access they grant is more considered. Swedish and Dutch consumers are more likely to simply leave than resign themselves to giving data access.

The goal isn’t a higher yes rate. It’s a more honest one.

Turning a resigned yes into a real one

Most consent advice is written to raise your opt-in rate. That’s the wrong goal – a higher rate is exactly what resigned consent already gives you. These checks are aimed at something narrower: making sure the yes you get is one people meant.

Consent experiences built this way are likely to generate lower opt-in rates than the one you have today. That’s the point. The yeses that remain are far more likely to hold up when something goes wrong.

Getting there isn’t a one-time fix, it’s worth building alongside people who’ve done it before.

Talk to a Usercentrics specialist about closing the resigned-consent gap in your own AI rollout.

Explore case studies from organizations turning privacy infrastructure into measurable performance.

The brands consumers will grant AI access to in 2027 are deciding how they ask right now.

11,000 consumers. 7 markets. Conducted by Sapio Research in March 2026. Accurate to ±0.9% at 95% confidence. Sweden is a new market this year; YoY comparisons exclude Sweden.

The State of Digital Trust Report is provided for general informational and research purposes only and does not constitute legal, compliance, regulatory, or professional advice of any kind. Any views, projections, or statements contained herein reflect opinion based on consumer research as of the date of publication and are not promises or guarantees of any commercial, regulatory, or other outcome. Readers should not rely on this report as a substitute for independent professional judgment. To the fullest extent permitted by applicable law, Usercentrics disclaims all liability for any loss, damage, or harm arising directly or indirectly from any action taken or omitted in reliance on the contents of this report.