4 Ways to Talk Honestly About AI in the Contact Center

If you’re not hyperventilating about AI yet, you’re just not getting it.

AI is a magical solution that can handle 100% of customer interactions flawlessly. AI will cut costs, reduce headcount, enable 24/7 service, improve the deflection rate, capture customer sentiment without ambiguity and improve productivity by 90% OR MORE.

But all this AI hype is causing real harm. If everyone else’s contact center AI project is going great, just so so great, you may be reluctant to confide that your AI project is going OK, just kind of OK.

And if we’re not candid with each other when our AI reality doesn’t match what a vendor promised us or what we promised our boss, we risk losing our community, which is one of the most precious things we have. Candor holds our community together.

Let’s be real with each other. Here’s how to talk honestly about AI in the contact center, so we can continue to rely on each other during this epic period of change.

1. Call each AI tool by its name. Your CEO may have charged you with “implementing AI,” but you know that the term “AI” refers to different tools with different functions. Rein in the AI hype by referring to specific AI tools and how you’re using them in your contact center. Maybe you’re using an AI tool for pre-call authentication and customer data capture or one that summarizes calls automatically. Instead of saying that your contact center is “using AI,” mention which tool you’re using and what it does.
2. Respond to your frontline agents’ fears of being replaced by AI tools. There’s no elephant in the room if you talk about the elephant in the room (or something like that). Using AI tools will affect headcount in your contact center, but lots of other conditions affect agents’ job security too: tariffs, taxes, or supply chain problems; remote vs in-office assignments; outsourcing; competition in the marketplace, etc. You’ve addressed these conditions in the past, so you have developed the skills and experience you need to talk to your agents about their career prospects. Show respect by talking candidly with agents about your contact center’s AI roadmap and its implications for their jobs. Help them identify the skills they’ll need as your organization adopts more AI tools and provide training, so they can develop those skills.
3. Tell your contact center’s AI story truthfully, especially when you’ve had to reconsider your original plan. Don’t imply that AI tools are providing better customer service when they’re just providing cheaper or faster service. Be like Klarna. In 2024, the Swedish “buy now, pay later” company replaced 700 of their approximately 3,000 customer service agents with chatbots and AI. While Klarna did save money, they found AI agents weren’t up to handling some complicated issues, such as identity theft. They had to rehire a few humans. If your AI project had an undesired outcome and you had to take a do-over or shut it down, talk about that. Tell your colleagues what you learned.
4. Acknowledge the effort of cleaning up your "dirty data.” Your AI tool can’t pull accurate information out of thin air. It relies on CRM data, CSAT, NPS scores, knowledge base articles, documentation, product specs and so much more. Much of this information is siloed, dirty, incomplete or poorly written. Be honest about the massive, ongoing effort required to clean up the information that AI tools train on, so they can understand customer intent, personalize responses and automate tasks. Like so many tasks in the contact center, using AI is less about the sparkly new tool than it is about the daily effort of running a clean operation. Tell the truth about the grind.

Our ICMI community means the world to me — personally and professionally. Let’s weather this AI hype storm by being real with each other. Half-truths about our AI implementations hurt us all. Using AI tools can be messy and complex; these tools require a substantial human shift. The most successful contact centers will be led by the most transparent leaders.