Customers are three times more likely to use a third-party generative AI tool like ChatGPT or Claude for customer service than a brand-owned chatbot, a Gartner survey of more than 3,500 B2B and B2C customers found. Third-party AI usage has doubled in the past year alone, while company-provided chatbot usage has not statistically increased since 2022 - a gap that should push customer support leaders to rethink where they're placing their AI bets.
"So a doubling of third-party AI, no increase in customer service chatbots - it definitely should give pause to leaders who are thinking that the future of customer service is in these company-owned AI chatbots," Eric Keller, senior director analyst in Gartner's customer service and support practice, told CX Dive.
Why customers default to third-party tools
Two-thirds of consumers now use generative AI in their personal life, work, or both, but most of that usage flows through tools they already know and trust. Keller said the preference reflects comfort and quality. "They use these tools in every aspect of their life, and they increasingly trust those tools and trust the responses that they get. So it makes a lot of sense that when they experience a customer service issue, they turn to this tool that they use every day rather than going to a chatbot on a company's website that they don't use often."
For support teams investing in AI for Customer Support, that means simply layering generative AI onto an underused chatbot is unlikely to boost engagement. "If customers are not already engaging with your chatbot, simply putting AI in that chatbot is probably not going to drive more engagement," Keller said. "So you really need to drive intentional adoption strategies."
Chatbots that answer but don't act
Customers aren't just asking questions - they want to complete tasks. Gartner found that 58% of customers have used generative AI to finish a task, and the number rises to nearly three-quarters among B2B customers. Yet many brand chatbots stop at providing information and then hand off the actual transaction with a link to another part of the website.
"That's where the brand's chatbot is uniquely positioned right now because you can't do that through ChatGPT," Keller said. "But often the brand's chatbot will just simply answer questions. And then if you want to transact, if you want to make an update, if you want to add something, if you want to change something, it's going to send you a link to go somewhere else on that website and do it. That's a missed opportunity."
Redesigning the front door
The familiar chatbot widget tucked into the bottom-right corner is starting to feel outdated, Keller said. Leading organizations are replacing it with a single conversational interface that shapes the entire digital experience - a search box that asks, "What are you trying to do with us today?" instead of a homepage full of menus and links. The shift turns a navigational experience into a conversational one, making it easier for customers to both get answers and take action in one place.
Why this matters for customer support leaders
The data doesn't say customers reject AI in service - they're already using it heavily, just not on your website. Support leaders need to align investment with actual customer behavior: prioritize tools that let customers complete tasks, not just read answers, and redesign the entry point so it feels familiar and capable. An AI Learning Path for Call Center Supervisors can help teams build the skills to move beyond a bolt-on chatbot and toward a unified, action-oriented support experience that matches what customers already expect from the AI they use every day.
Your membership also unlocks:
