Centrica CEO Chris O'Shea backs AI as 1,300 call-centre jobs are axed

Centrica, the FTSE 100 owner of British Gas, is preparing to axe 1,300 call centre jobs, and its CEO, Chris O'Shea, says customers would rather talk to an AI chatbot than staff. That claim sits at the center of the company’s defense of the cuts, even as Centrica reports a rise in retail profits and signals it is chasing “bigger margins.”

The plan is not a single one-off layoff. Centrica says it will cut 800 jobs as part of a “targeted deployment of AI tools,” and that is on top of the 500 cuts it confirmed last month. So the headline number becomes hard math for anyone watching operational strategy: 800 now, 500 previously, 1,300 total, with AI as the rationale and profit momentum as the backdrop.

This is a story about incentives, not just technology. In retail energy, margins are the difference between stable earnings and a year that feels like it was written by a stress test. When a company reports rising retail profits while simultaneously shrinking its customer support workforce, the message to markets is that the cost base can be reshaped without waiting for demand to grow. The “bigger margins” framing matters because it suggests this is not only about modernizing service, it is about redesigning the economics of how households get help paying bills, changing plans, or resolving issues.

It is also a story about customer expectations colliding with operational reality. Centrica’s CEO is effectively arguing that the locus of service demand is changing, with “most households” preferring an AI chatbot over dealing with company staff. That kind of statement is consequential because it attempts to de-risk the human side of workforce cuts by shifting the burden of choice to customers. Whether or not every household agrees, the logic is clear: if customers prefer faster, always-on automated answers, then maintaining large phone staffing levels becomes harder to justify.

For executives, the key second-order question is what happens when the “preference” claim meets edge cases. AI chatbots can handle routine queries well, but customer service in utilities also includes complaints, exceptions, billing disputes, and situations where escalation is not optional. If the company’s AI deployment is truly targeted, it implies a split in the workflow: automation for broad, repetitive requests, human escalation for the messy tail. That transition is typically where boards get nervous, because service levels, complaint rates, and reputational risk can move quickly when you reduce human capacity.

This is where the regulatory backdrop becomes relevant, even if the source does not cite specific regulators. UK energy suppliers operate in a heavily scrutinized environment, and customer treatment is a recurring policy concern. When a company changes how customers access support, regulators and consumer advocates tend to focus on fairness, transparency, and when human help is available. The business case is clear. The compliance and operational details are where projects can slow down or get contested.

There’s also a capital market angle. Centrica is in the FTSE 100, and the combination of rising retail profits plus layoffs framed as AI-driven efficiency is a signal that management wants the market to see this as modernization with upside. The risk is that any mismatch between the AI promise and customer experience could show up in harder-to-ignore metrics later, like retention, churn, or complaints. Timing matters too. The company is stacking 500 cuts confirmed last month on top of 800 tied to AI tools, which suggests the transition is not waiting for an extended pilot cycle.

Peers across utilities and other consumer subscription businesses should read this as a warning and a blueprint at the same time. The warning: workforce reductions can accelerate quickly when margins are the priority and leadership has a crisp automation narrative. The blueprint: management will likely try to tie AI deployments directly to cost actions, then use profit reporting to bolster credibility. For boards and CFOs, the strategic stake is straightforward. If customer-facing automation becomes a margin lever, the companies that govern it well will protect service quality and regulatory alignment, while the ones that govern it loosely will discover that customer support is also brand support.