Your AI Rollout Might Be Undermining Customer Experience

The Gist

• Start with customers. AI should support CX goals like ease, trust and personalization, not just operational KPIs.

• Tech alone fails. AI can’t fix broken customer journeys without deeper redesign.

• Agentic AI risks. Newer AI tools can act on their own, which raises new challenges for safety and oversight.

In the fall of 2022, OpenAI released ChatGPT, and everything changed. Witnessing the power and newfound accessibility of generative artificial intelligence, suddenly every CEO was pushing to implement AI. As business teams looked for AI opportunities, customer experience (CX) and the contact center emerged as areas particularly ripe for AI interventions.

Of course, people in the CX industry had already been using conversational AI, machine learning, natural language processing and other advanced technological innovations for years. But generative AI at this level was new, so we sought to understand its implications, separate fact from fiction and find ways to prove its ROI.

The last two years have been a journey, and we’ve learned a lot about what’s possible, practical and profitable. This is still a rapidly developing landscape, but the core principles of great CX remain the same, even in the era of AI.

Below, I’ll share the truths that remain, the new ways AI is affecting customer service and where we see change coming.

Table of Contents

What Still Matters in Customer Experience

While AI is disrupting how customer experiences are delivered, the core principles of what customers expect remain the same. Customers want brands to make experiences effortless, anticipate needs, personalize interactions and instill trust.

While AI can allow these principles, customers don’t care about the technology itself, and whether or how it is delivered by AI. They only care that their experience is faster, easier and more personalized.

This is where many companies get it wrong. They start on the inside, focusing on where AI can reduce cost, increase revenue and improve operational KPIs. They believe that optimizing these metrics will naturally lead to better customer satisfaction. That logic is flawed.

Instead, companies need to start on the outside with the customer and focus on making their experience effortless, anticipating their needs, personalizing their experiences and instilling trust. These actions will have an amplified impact on the key measures of cost, revenue and operations just the same. This is the starting point, and AI can help achieve these goals. But companies must start from the outside and work their way in.

Related Article: Mastering Personalized Customer Experience for Growth

Traditional CX Problems Persist

Companies have long looked for ways to transform customer service from a cost center to a strategic asset, or even a profit center. Today, as in the past, many companies strive to do this by looking for opportunities to automate repetitive activities and high-volume/low-complexity interactions.

While AI can automate at a higher level than its predecessors, technology alone cannot improve customer experience. Often, processes and systems need to be strategically redesigned or even scrapped altogether.

Take the traditional car buying process, for example. At most dealerships, customers still spend hours filling out paperwork and waiting for approvals before they finally get the keys. It’s a process in desperate need of reinvention; no mere technology upgrade or AI band-aid will fix it.

AI Is Taking Over

While the same CX challenges and goals remain, AI in 2025 is now playing a significant role in customer experience. AI is disrupting every CX touchpoint, with its impact extending beyond automation to fundamentally redefine customer journeys and drive business growth.

The convergence of contact center as a service (CCaaS), customer relationship management (CRM), communications platform as a service (CPaaS) and digital CX is leading to a new view of customer experience as a service (CXaaS), with AI as the driving force.

Examples of AI-Enabled CXaaS in Action

Editor’s note: This table highlights key ways artificial intelligence is transforming customer experience as a service (CXaaS).

Power and Risk Ahead for AI in CX

That agentic AI is the “new big thing” isn’t exactly a hot take. Gartner is already predicting that agentic AI will autonomously resolve 80% of common customer service issues by 2029.

For those not familiar with agentic AI, here’s a quick primer. Unlike generative AI, which excels at generating new content across formats, agentic AI is action oriented. It goes beyond content creation to empower autonomous systems capable of independent decision making and actions. These systems can analyze situations, formulate strategies and execute actions to achieve specific goals, all with minimal human interaction.

Pretty cool, right? Well, yes … until it goes wrong. Take the case of the Washington Post reporter who, with the help of agentic AI, accidentally bought a dozen eggs for more than $30.

That example is an annoying and expensive inconvenience, but think about the harm an autonomous AI agent could cause a brand or a customer. Generative AI might give customers the wrong information (as in the case of the Air Canada chatbot), but agentic AI can actually do the wrong thing.

As with all forms of AI and machine learning, agentic AI will require guardrails and human oversight to avoid harmful customer experiences, privacy violations, financial losses and other risks. Organizations adopting it must continually refine their AI governance procedures to remain aligned with global regulations.

Related Article: A Practical Guide to AI Governance and Embedding Ethics in AI Solutions

The Work That Remains

If recent history has taught us anything, in 2025 we can expect to see incremental AI advances and possibly more big disruptions on the level of DeepSeek. However, in my opinion, anticipating the next innovation is less useful than learning lessons from the AI we already have.

First, we’ve clearly seen that AI can cut costs; next we need to focus on balancing cost cutting with meaningful customer experience improvements. AI is ready to drive customer engagement and revenue.

The second lesson is that it’s time for businesses to move beyond AI experimentation and develop trusted, governed AI strategies. 

Finally, we’ve learned that AI in CX isn’t a future state. It’s already here, and businesses that differentiate with AI will have a competitive advantage.