Your customers would rather use ChatGPT than your chatbot, Gartner finds

Gartner surveys of 3,566 customers and 1,303 senior leaders found customers are three times more likely to use third-party GenAI tools like ChatGPT than company chatbots for customer service, while only 24% of customer-service AI investments yielded positive financial returns. The research highlights a misalignment between enterprise AI deployments and customer expectations for task-completion rather than just answers.

Businesses spent big on AI to answer their customers. The customers had other ideas. People are approximately three times more likely to use a third-party GenAI tool, such as ChatGPT, than a company’s own chatbot to sort out a service problem, according to Gartner, Inc. https://www.gartner.com/en/newsroom/press-releases/2026-07-08-gartner-survey-finds-customers-are-three-times-more-likely-to-use-third-party-genai-than-company-provided-chatbots-for-customer-service The firm, a research and advisory company, surveyed 3,566 customers in February and March. Use of outside GenAI tools during service interactions has nearly doubled in a year. Use of company chatbots has barely moved since 2022. Consumers, meanwhile, have embraced these tools, and half the US public https://thenextweb.com/news/pew-research-americans-ai-chatbot-skepticism-regulation-distrust-2026 now uses AI chatbots. Money in, little out The spending tells the other half of the story. A separate Gartner survey of 1,303 senior leaders found customer-service teams put a median 12 per cent of their 2025 budget into AI, more than any other business function. Only 24 per cent saw a positive financial return. “The disappointing impact of customer-facing GenAI investments has less to do with technology limitations and more to do with misalignment with customer expectations,” said Eric Keller, a senior director analyst at Gartner. He argued firms should build AI-enabled service journeys across digital and voice, not standalone bots. Customers want action, not answers The behaviour has shifted in a second way. People no longer just want an AI to answer a question. They want it to do the task. Among GenAI users, 58 per cent said they had used it to act on their behalf. In business-to-business settings that rose to 74 per cent. Booking an appointment, submitting a document, updating an account: customers expect the bot to finish the job. Most company chatbots still stop at answers. That gap helps explain why people drift to tools that already act on their behalf https://thenextweb.com/news/ai-shopping-prime-day-amazon-chatbots . Why it matters Customer service has become an early test of whether enterprise AI pays. So far the money flows in faster than the returns, an adoption-versus-impact gap https://thenextweb.com/news/ai-adoption-divide-safety-critical-dan-macdonald seen across industries. Meanwhile the AI labs circle the same work, and rivals keep buying up support-AI https://thenextweb.com/news/salesforce-acquires-fin-intercom-3-6-billion . Companies that treat GenAI as a bolt-on chatbot may keep losing their own customers to someone else’s AI https://thenextweb.com/news/openai-northslope-acquisition-enterprise-ai-deployment . One more thing: people still want a human on the line when the bot fails. Get the TNW newsletter Get the most important tech news in your inbox each week.