When AI personalization backfires in customer experience
Personalization used to be the gold standard of modern customer experience. By using data to anticipate needs, you can make every interaction feel tailored. This signals care, right?
In reality, it’s starting to signal surveillance.
When we surveyed 600 U.S. consumers in October 2025, we found a growing tension between what companies think personalization means and what customers are actually comfortable with.
Only:
- 17% said they are very comfortable with companies using their personal data to personalize support
- 46% said they are somewhat comfortable
- 37% said they are not comfortable
That means more than a third of customers feel uneasy about how their data is being used in support experiences.
That’s not a personalization advantage. That’s a trust warning.
When helpful turns into intrusive
There’s a thin line between “You understand me” and “You’re watching me.”
Customers described situations where personalization crossed that line:
- Referencing old support issues they didn’t expect to revisit
- Mentioning history that felt private
- Proactively surfacing sensitive account details unprompted
- Over-scripted “personalized” responses that felt robotic
What companies see as intelligent automation, customers sometimes experience as exposure.
When systems surface every piece of data available instead of just the context needed, the experience feels invasive rather than supportive.
The expectation gap
There’s a disconnect between leadership enthusiasm and customer comfort.
Industry surveys show that CX leaders overwhelmingly view customer data as strategic for personalization. Many AI roadmaps prioritize predictive engagement and proactive service as differentiators.
Customers are more cautious.
Our 2025 survey data shows:
- 75% prefer human-first support
- Nearly 90% show reduced loyalty if human support is eliminated
- 34% say AI support makes things harder
Layer aggressive personalization on top of shaky trust in AI, and skepticism grows.
Customers don’t want hyper-targeted nudges. They want relevant help. There’s a difference.
Why over-personalization erodes trust
Trust is built through transparency and control.
When personalization feels invisible or unexplained, customers feel manipulated. When it feels intentional and bounded, it feels helpful.
Consider two scenarios:
Scenario one
A support agent says, “I see you contacted us about a billing issue in March. Let’s pick up from there.”
This feels continuous and supportive.
Scenario two
A chatbot says, “We noticed you viewed our refund policy three times this week. Are you unhappy with your purchase?”
This feels intrusive.
Both use data, but only one builds trust.
Personalization fails when it centers what the company knows instead of what the customer needs at that moment.
The personalization reset in 2026
Leading CX teams are shifting from hyper-personalization to intentional personalization.
That means:
- Surfacing only the context necessary to resolve the issue
- Giving customers clear visibility into what data is being used and why
- Making it easy to revisit prior tickets or preferred channels
- Using AI to guide, not predict aggressively
Instead of mining every behavioral signal, teams are asking a simpler question: What does this customer need right now?
Personalization without trust doesn’t scale
AI-powered personalization was supposed to increase loyalty. But loyalty doesn’t come from knowing everything about a customer. It comes from respecting them.
Customers want:
- Resolution
- Empathy
- Continuity
- Choice
They don’t want to feel profiled.
As AI capabilities expand, the brands that win in 2026 will be those that use personalization to reduce effort and increase clarity without crossing into surveillance territory.
Because once trust breaks, no amount of data sophistication can repair it.
Download the full 2026 trends report
This blog is based on insights from Glance’s Customer Experience 2026 Trends Report, built from a survey of 600 U.S. consumers conducted in October 2025.
The full report explores:
- Why AI-driven personalization is creating a trust gap
- What customers are actually comfortable sharing
- How to design personalization that builds loyalty instead of skepticism
- The Reset, Rehumanize, Refocus framework for 2026
Download the full report to explore the data and strategy shaping CX in 2026.
Personalization in customer experience refers to tailoring support interactions based on customer data, history, preferences, or behavior.
Customers feel uncomfortable with personalization because overuse of behavioral data can feel invasive. In our survey, 37% of customers said they are not comfortable with companies using personal data for personalization.
No, AI-driven personalization isn’t inherently bad for customer experience. When used thoughtfully and transparently, AI can improve relevance and reduce effort. The issue arises when personalization feels excessive or manipulative.
Companies can personalize without violating trust by limiting data use to what’s necessary, being transparent about data usage, and allowing customers control over their preferences.
Yes, customers prefer human support over AI. In our survey, 75% said they prefer human-first support, and nearly 90% showed reduced loyalty if human support is removed.
All statistics cited come from Glance’s Customer Support Survey of 600 U.S. consumers conducted in October 2025, featured in the 2026 trends report.
