AI is supposed to drive CX improvements, but gaps in confidence about its capabilities are keeping companies from making the most of it. Although AI can analyze data and deliver real-time CX insights, many organizations avoid using it for customer-facing interactions. They worry about breaking processes or frustrating customers with subpar results. Closing these confidence gaps can help companies improve their CX at a time when it seems to be getting worse.
US consumers’ views of major-brand CX have declined over the past four years, hitting anall-time low in 2025, according to Forrester Research. “Disappointing implementations of potentially game-changing technology (including AI)” were cited by Forrester as factors in the trend. Transforming CX will require using AI in ways that generate reliable outcomes. The first and most important step is building internal use cases and rigorously refining them before adding customer-facing use cases.
Understanding the AI trust gaps in CX
It’s not just organizations who are worried about how well AI will work in customer-facing use cases. Consumers don’t trust AI with certain tasks either, especially complex ones. YouGov found that while 65% of Americans trust AI to compare prices among stores, only 35% trust AI with customer service inquiries. That may arise from early chatbot and IVR experiences that didn’t make things easier. Now, companies need to deliver more intuitive and useful AI-backed experiences to overcome that trust gap.
Companies at the leading edge of customer-facing AI will create a new gap. Once consumers experience how much easier AI agents can make their experiences, they’ll be less willing to engage with brands that don’t offer similar experiences. Why would a busy customer spend time and mental energy clicking through multiple steps to compare hotels and book a stay if another hospitality brand’s AI agent can generate a list of available properties based on their destination, dates, budget, and other criteria?
Evolving from AI navigation to reliable AI conversation
To close the gap between CX expectations and reality, companies must bridge the distance between how they use AI now and how they’ll deploy it in customer-facing scenarios. That process starts with understanding the shift from AI as a self-service navigation tool to a self-service conversation partner.
As consumers get more accustomed to having conversations with their personal AI assistants to find what they need, they expect similar conversational experiences from brands. Using the hotel example again, customers expect more than a smoother way to search listings. They’d rather ask a question and get a useful answer.
Customers also expect service options initiated by the brand. For example, imagine you’re clicking around on a gift website and the site’s AI assistant says, “It looks like you’re searching for a graduation present. Here are our most popular options.” This can jump-start a productive conversational search.
As these kinds of AI-powered experiences become more common, channel journeys will give way to agent-led journeys on websites, in apps, and personal AI assistants. As leading-edge companies use AI in these ways to transform their CX, their competitors will have to keep up or lose out.
Start building internal AI use cases to strengthen trust
Keeping up with CX transformation starts with internal use cases that help teams learn and build confidence in their ability to leverage AI. To build a strong initial use case, separate AI reality from hype. Invest time upfront to learn data, integrations, and functionality requirements so you can accurately map your use case and set KPIs. These internal use cases also cultivate trust, which drives adoption.
For example, trust became a governance question as a global telecom company began scaling AI-powered employee experiences. The solution was to establish a clear framework for how AI agents are created, maintained, and measured to ensure consistency as adoption grew. Prior to implementation of this framework, agentic experiences were enabled by integrations with two key platforms. Today there is a vetted, prioritized pipeline of nearly 50 tools that stakeholders can integrate for better EX.
This positions AI as a managed, accountable product ecosystem that employees feel confident using, rather than as a standalone tool. This is the kind of use case that can provide learnings for customer-facing AI solutions.
Using AI to support contact center employees can be a good AI pilot. It allows your organization to leverage customer data for training employees and solving customer queries faster and more accurately. For example, when a customer calls in with a product issue, an AI agent can help the customer service representative quickly find the right information to troubleshoot or recommend.
When the support agent consistently delivers accurate information and effective support, it’s time to explore customer-facing agentic micro-interactions. As with the internal use case, the idea is to start small to build trust and confidence. For example, some retailers have AI agents on their product pages, like Home Depot’s “Ask about this product” feature, that prompt customers to ask questions and learn more about the product’s features.
For example, a customer looking at an outdoor furniture set might ask the agent about how to take care of the set and how long it’s guaranteed to last. The customer still has access to all the product details on the page, but the agent saves time by surfacing the information that matters most to each shopper.
Asking and answering these questions drives engagement and helps customers feel more confident about their purchase. At the same time, limiting the AI agent’s scope to product information makes it easier to control output quality.
Start small, but start closing the AI CX gap now
Closing the AI‑trust gap in CX is urgent, but it’s not a sprint. Start with tightly scoped internal pilots and build on what you learn. As your organization develops fluency with agentic AI in one area, extend that knowledge into small customer-facing use cases. As you earn customer trust, you also generate data to help extend and scale agentic use cases across the customer journey, moving from AI-assisted navigation to AI-powered conversations that give customers the experience they want.
Coauthor
Emily Lesinski has 15 years of experience in digital analytics with a focus on the marketing and advertising industry in both the US and UK. In addition to digital analytics, she also has experience in content creation, editing and strategy to help clients maximize their data-driven outcomes.