Quick Summary
-
In their fervor to quickly capitalize on AI-powered customer service, companies have rolled out flawed solutions that are eroding customers’ trust in the technology.
-
Many companies approach AI-based service enhancements too narrowly, using it mainly to make existing customer touchpoints faster, cheaper, or more efficient.
-
The overlooked opportunity is using AI to strengthen customer engagement by introducing new, meaningful touchpoints that couldn’t practically exist without the technology.
Finally, there’s an AI customer service implementation worth emulating.
Such examples have been difficult to find. In their fervor to capitalize on the new technology, companies have been rolling out half-baked AI-powered service solutions – the kind that trigger customer rage rather than raves.
You’ve no doubt encountered these types of automated service platforms: AI chatbots that repeatedly misunderstand your requests, can’t find answers to your questions, and make it nearly impossible to connect you with a live representative.
It should therefore come as no surprise that a majority of consumers (54%) say they find AI customer service frustrating (per Five9’s 2026 Consumer CX Report). The same study found that 4 out of 10 consumers are less likely to patronize a company if it uses AI for customer service.
Are there scenarios where AI can make existing customer service interactions better? Sure, and as the technology improves, those use cases will no doubt expand.
The problem, however, is that many companies approach AI customer service enhancements through too narrow a lens. They focus on the improvement of existing customer touchpoints, as opposed to the creation of entirely new ones.
As a result, typical AI implementations center on making existing customer interactions faster or cheaper – for example, by providing automated answers to common questions, or obviating the need for customers to navigate through hellish 800-line menus. And while those efforts are noble and should continue, they fail to address another important strategic consideration: How could we use AI to engage our customer in meaningful ways that weren’t previously possible?
“How could we use AI to engage our customer in meaningful ways that weren’t previously possible?”
It was that line of thinking that led AT&T to launch AI-generated summaries of customer calls – and it’s one of the more admirable applications of customer service AI out there.
Once opted in to the feature, customers who call AT&T (be it with questions, requests, or complaints) receive a text message recapping the conversation – what was discussed, what actions were taken, and what comes next. This simple innovation improves the customer experience in several ways (all very much aligned with Watermark’s psychology-based 12 Principle framework for creating loyalty-enhancing experiences):
- It makes the experience more effortless. Presuming the AI-generated recap is complete and accurate, it saves customers from having to take notes or rely on memory to keep track of what was discussed and what the service representative promised to do. AT&T takes responsibility for documenting the interaction, rather than leaving that work to the customer. Why this matters: A low-effort customer experience cultivates loyalty, while a high-effort one erodes it.
- It gives customers a greater sense of control. By documenting actions taken and commitments made, the text recap mitigates the kind of ambiguity that customers despise (“What did the service rep really promise to do and by when?”). That, in turn, helps the customer feel more in control of the interaction. Why this matters: People feel worse about experiences that rob them of a sense of control, and any uncertainty that exists about “what comes next” feeds into that discomfort.
- It finishes the call interaction on a higher note. Rather than ending with the typical call closing courtesies (“Is there anything else I can help you with today?”), the interaction concludes with something that is pleasantly surprising and genuinely useful – a written record of what happened and what comes next. Why this matters: The last part of an interaction exerts a disproportionate influence on people’s perception and memory of the whole encounter. By finishing strong with the text recap, customers will tend to view the entire call experience more favorably than would otherwise be the case.
If a simple, written recap of a phone call could have such a material impact on the customer experience, then why don’t more companies do it? And the answer is – it’s tough to scale.
When your service representatives are fielding dozens of customer inquiries each day, carving out time for them to create text summaries of each call would have enormous implications for a contact center’s staffing and cost structure. It’s just not feasible to do, except perhaps in the smallest of businesses.
But AI changes the equation. An experience enhancement that was previously impractical, suddenly becomes quite viable.
So, if you’re looking for ways to leverage AI technology in customer service, don’t limit yourself to making existing touchpoints more effective and efficient. Think of AI as a vehicle for creating meaningful additional interactions with customers that wouldn’t otherwise be possible.
Because the best customer experience strategies don’t just focus on today’s touchpoints – they also invent tomorrow’s.
Disclosure: AT&T is a former client of Watermark. This article is based solely on publicly available information.