The human element remains pivotal in AI-driven customer service strategies
Customer service roles have long been viewed as particularly vulnerable to the rapid deployment of increasingly sophisticated artificial intelligence tools—technology that a senior executive at one of the world’s largest insurers branded as a “superpower.”
Germany’s Allianz has actively integrated AI-powered solutions, including a voice assistant launched in September that offers roadside support for breakdowns. Operating in over 20 languages, this system prioritizes the most urgent calls—such as a lone parent stranded in the middle of the night with a car inoperable—and immediately routes them to human agents.
In July, Allianz’s Australian arm introduced an app designed to speed up homeowners’ insurance claims for food spoiled during power outages caused by severe weather. “Generative and agentic AI is like a superpower we use to improve our customers’ lives,” says Josef Teglas, Allianz’s group head of data and AI. “We can cut turnaround times, be there when needed, and resolve problems and claims much more quickly.”
Yet Teglas stresses the need to roll out AI in a responsible and safe manner. Analysts interpret this prudence as reflecting a common concern: AI still struggles with complex, real-world customer situations, suggesting that call centers won’t soon become obsolete.
A recent Gartner report concluded that a fully automated customer service function is “both unlikely and undesirable.” The findings contrast with OpenAI CEO Sam Altman’s earlier warnings that widespread AI adoption could render certain job categories, including customer service, “totally, totally gone.”
The study, titled Agentless Customer Service Should Not Be Your Goal, predicts that by 2027, half of the organizations hoping to dramatically shrink their service workforce through AI will abandon those plans.
“AI and automation are transforming how customer service teams operate, but human agents remain indispensable for handling nuanced situations and fostering lasting relationships,” says Kathy Ross, a co-author of the Gartner study.
She envisions a hybrid future. “When customers require a human connection or when issues are complex or sensitive, AI is likely to act as a teammate to elevate service quality,” she explains.
Jo Causon, chief executive of the London-based Institute of Customer Service, also supports a blended approach to AI-powered customer service. “AI should not be seen as a complete substitute for human skills. It’s a tool that enhances human effectiveness,” she notes.
Expedia offers a practical example: chief product officer Shilpa Ranganathan states that AI handles more than half of customer inquiries, while staff take over for more intricate travel itineraries. “We use AI as the initial touchpoint, and the information it gathers is seamlessly handed to the human agent,” she says.
Jonathan Corbin, CEO of MavenAGI, which builds AI agents, agrees that AI is unlikely to fully replace customer service staff. “It’s not about replacing humans; it’s about empowering people to engage with customers who need a human touch the most—think of someone like my mom.”
John Leighton, EasyJet’s customer service director, explains that the airline uses AI alongside staff. The technology suggests responses to inquiries via phone, live chat, and email, which staff review and adjust before sending. He frames AI as an investment in customer experience rather than a pure cost-cutting measure.
EasyJet also employs AI-powered social media monitoring to identify customers needing urgent support. Rolled out in March 2023, the tool helped during Greece’s serious wildfires by flagging terms like “stranded” and “elderly,” enabling staff to reach vulnerable travelers promptly.
Nick McBrien, chief executive of The Very Group (UK-based online retailer), recalls initial staff apprehension about AI. That hesitation faded once employees realized the technology was designed to augment their performance. He points to the company’s record high customer-satisfaction results in the year ending June 2025, which he partly attributes to the guidance and information AI provides to employees, reinforcing the message that AI is there to support, not replace, workers.
