2025 was the year OpenAI CEO Sam Altman predicted that customer service jobs will become obsolete.
“I’m confident that a lot of current customer support that happens over a phone or computer, those people will lose their jobs, and that’ll be better done by an AI,” Altman said during a September interview on “The Tucker Carlson Show.”
But for all the bold talk of AI revolutionizing customer service and customer experience more broadly, contact centers are still staffed, customers still prefer humans for support interactions, and AI regulation here and abroad is muddying the water for deployment in CX.
The technology has advanced rapidly, but hurdles, from messy data to lack of adoption, have muted some pronouncements. Still, it’s easy to wonder: How will businesses use AI in customer experience in 2026?
CX Dive spoke to three experts for their predictions on the year ahead.
Most brands will use AI for discreet tasks, not agentic solutions
Agentic AI might be the talk of the town, but most companies will not be rolling out agentic AI solutions for customer experience in 2026, according to Isabelle Zdatny, head of thought leadership at Qualtrics XM Institute. Instead, they’ll use the technology for specific tasks.
“CX teams will continue to use discrete AI applications to solve specific challenges like applying natural language processing to unstructured data,” Zdatny said.
CX leaders will use AI for such tasks as building predictive models, like synthetic NPS, which allows them to reach out to customers who might be at risk of leaving.
“What I don't think we are going to see as much is those agentic AI systems that are orchestrating end-to-end processes and workflows,” she said. “In order to successfully deploy agentic systems, companies need to do the very hard, unglamorous work of actually mapping out their processes and workflows, making sure their data is clean.”
But organizations don’t work like sleek machines.
“They are built on top of a lot of undocumented and messy processes where, Susan in procurement has this workaround,” Zdatny said. “When companies try to apply AI on top of that, it just amplifies the dysfunction. It doesn't fix anything.”
Some companies will deploy AI too quickly and sabotage their self-service experience
Among those that do roll out a bot powered by "modern AI" — whether that’s generative, agentic, or a LLM — slightly more of these efforts will fail than succeed, according to Forrester Principal Analyst Max Ball.
While business interest in AI self-service applications have skyrocketed, Ball predicts about one-third of brands will roll out AI in self-service and fail. These businesses will have pushed AI solutions out before they are ready, most likely due to cost pressures.
“The root of most failures in customer self-service is the cost pressures that contact centers are constantly fighting," he said. "For decades, this pressure has led many contact centers to focus on self-service as a cost savings measure.”
A conversation handled in self-service costs around one-tenth of what an agent assisted call costs, he said. Brands that jump on the AI bandwagon too fast and try to do too much with self-service will run the risk of failing to resolve issues and harming their customer relationships.
“AI is ready for the right use cases, but even for those use cases, brands need to be ready before they deploy,” Ball said. “Data quality, good application design to ensure a bot is useful, and proper integration for escalation to an agent are just a few of the things that need to be mastered before an AI deployment can succeed.”
Even among brands that succeed, AI may cost businesses long-term loyalty. Some brands will lose loyalty in the year ahead because they fail to make real, emotional connections with human employees.
“Customers might still interact with you if you're the speediest and most convenient, but the second you know someone else offers a cheaper price, or someone else, allows two clicks instead of three clicks, there's nothing keeping those customers with you,” Zdatny said. “You're automating away your long term loyalty.”
Federal AI directives will speed up AI deployments, but at a cost
Earlier this month, President Donald Trump signed an executive order to challenge and eliminate a patchwork of state AI laws. Currently, a number of states, including California, Utah and Colorado, have their own AI laws regulating how businesses use AI with consumers.
The federal directive will accelerate brands deployment of AI in customer experience, but it will come at a cost, according to Dan Hartman, director of CX outbound product management at CSG.
“For businesses, a unified national standard removes friction and speeds up adoption,” Hartman said. “But many state-level laws exist to protect consumers from muddy data use, discrimination and misleading automation. When those guardrails fade, brands take on a bigger CX risk.”
Brands will have more freedom to use AI in high-stake moments, from pricing to agentic systems that act on a customer’s behalf. While this innovation will move faster, experiences will appear less friendly to customers, Hartman says. “Ultimately, that falls back on the brand. Customers will blame the brand for the added friction, not the policy.”
The brands that will avoid this blowback will take consumer protection into their own hands.
“They’ll invest in governance, accountability and intentional limits on automation,” Hartman said. “Long-term, the winners will be the brands that keep their eye on the ball in CX and don’t get distracted by short-term cost-cutting wins.”
