Centrica, the parent company of British Gas, has announced plans to eliminate 1,300 call centre positions over the next two years. The decision is driven by what chief executive Chris O’Shea describes as a fundamental shift in customer behavior. Of these cuts, 800 roles will be removed as part of a "targeted deployment of AI tools," adding to the 500 job cuts already announced last month. Customer service teams in Glasgow, Edinburgh, Cardiff, Leicester, Stockport, and Leeds will be affected as the company accelerates its transition to digital-first support.
The job reductions will occur through a combination of natural attrition—positions left unfilled when employees leave—and compulsory redundancies. Trade unions have reacted strongly, warning that Centrica's investment in artificial intelligence effectively hands hundreds of human jobs over to chatbots. The union Unite has called for negotiations to minimize the impact on workers and to ensure that AI deployment is accompanied by adequate retraining and redeployment opportunities.
O’Shea has pushed back against the narrative that AI is the primary driver of the cuts. He notes that more than 90% of British Gas customers now begin their service interactions through digital channels such as the company's app or website. In addition, call volumes have declined by 20% over recent years. Using an app or website first, however, does not necessarily mean customers prefer speaking to a chatbot over a human agent. This distinction lies at the heart of the debate over the role of AI in customer service.
Do Customers Actually Prefer AI Support?
Several recent surveys paint a complex picture of customer attitudes toward AI in customer service. A February 2026 SurveyMonkey report found that 79% of Americans strongly preferred dealing with a human support agent, while only 8% actively preferred AI. Furthermore, 81% of respondents believed that companies introduce AI primarily to save money rather than to improve the quality of customer service. These findings suggest that there is a significant gap between corporate enthusiasm for AI and consumer trust in it.
Another study, conducted by Pegasystems in partnership with YouGov, surveyed more than 4,700 consumers across North America and the UK. The results showed that 66% of participants preferred human-led support. Nearly half (46%) reported that AI rarely or never solved their problem, and only 2% wanted to interact exclusively with chatbots. While these surveys do not focus specifically on British Gas customers, they indicate a strong general preference for human interaction, especially when issues are complex or emotionally charged.
Centrica's O’Shea may be conflating a preference for digital tools—such as checking a bill or scheduling an appointment online—with a preference for AI. Many consumers appreciate the convenience of self-service portals for routine tasks but still want the option to speak to a trained human when a problem escalates. The distinction is critical: digital channels often include human-assisted options, whereas AI-only systems can leave customers feeling frustrated and unheard.
When Does AI Customer Service Work Better?
AI can be highly effective for straightforward, routine inquiries where speed is paramount. Research published last year cited an Intercom survey in which 61% of consumers said they preferred AI when they needed a fast answer to a simple question—such as resetting a password or checking account balance. For these tasks, chatbots can provide instant responses 24/7, reducing wait times and improving efficiency.
The financial incentives for companies to adopt AI are substantial. One industry report estimated that AI can reduce the average cost of a customer interaction from $4.60 to $1.45—a savings of nearly 70%. Moreover, AI agents can resolve between 76% and 92% of basic ecommerce requests without human intervention. For a company like Centrica, which handles millions of customer interactions each year, the cost savings are significant enough to justify major investments in AI infrastructure.
But cost savings are only part of the equation. Customer satisfaction metrics often decline when AI is used for complex or sensitive issues. A billing dispute, a service outage, or a complaint about a vulnerable customer's treatment are scenarios where empathy, nuance, and human judgment are critical. In such cases, AI can struggle to understand context or offer appropriate solutions, leading to customer frustration and churn.
The deployment of AI in customer service also raises questions about job quality and the future of work. Call centre roles have long been a gateway to employment for many people, offering relatively stable jobs with training opportunities. As AI takes over more routine tasks, workers may need to upskill to handle higher-value interactions that require emotional intelligence and problem-solving skills. Centrica has stated that it will offer retraining programs, but unions remain skeptical about the adequacy of these measures.
Beyond Centrica, many other companies across industries are grappling with similar trade-offs. Banks, telecoms, and retailers are all investing heavily in AI customer service tools, often accompanied by workforce reductions. The trend is global: a 2025 report from Gartner predicted that by 2027, one in four companies would have replaced at least half of their customer service agents with AI. Yet consumer backlash remains a persistent risk. When airlines or insurance companies have eliminated human support options, they have faced public criticism and regulatory scrutiny.
The key to successful AI adoption lies in thoughtful design and transparency. Customers are more accepting of AI when they know they are interacting with a bot and can easily escalate to a human if needed. Companies that hide the fact that customers are talking to a machine—or that make it difficult to reach a human—often erode trust. The best practice is to use AI for initial triage and simple tasks, while routing complex or emotional issues to trained human agents.
For British Gas customers, the experience may vary depending on the nature of their inquiry. Checking a bill or changing an appointment online is likely to be handled seamlessly by AI. But a serious billing dispute or a report of a gas leak still demands human expertise. The available evidence suggests that most people want the option of speaking to a person when the problem becomes complicated. Centrica's job cuts, therefore, carry a risk: if customers feel their needs are not being met, they may take their business elsewhere.
In a competitive energy market, customer service can be a differentiator. While AI can improve efficiency and reduce costs, it cannot replace the trust and empathy that human agents provide. The challenge for Centrica—and for every company embracing AI—is to strike the right balance between automation and human touch.
Source:Digital Trends News