AI and energy: the tensions reshaping UK innovation in 2026

AI’s quiet revolution in Britain’s energy sector

The UK’s energy industry is undergoing a seismic shift—one driven not by wind turbines or nuclear reactors, but by artificial intelligence. Centrica, the parent company of British Gas, announced this week that it will cut 1,300 call centre jobs, citing customer preference for AI chatbots over human interaction. CEO Chris O’Shea framed the decision as a response to consumer demand, telling investors that “most households would rather speak with an AI chatbot than deal with the company’s staff.” The move, which follows 500 job cuts announced last month, reflects a broader trend: the automation of customer service roles in utilities, a sector long considered a bastion of stable employment.

The implications extend beyond Centrica. Ofgem, the UK’s energy regulator, has yet to comment on the job losses, but the shift raises questions about the future of work in an industry already grappling with the transition to renewables. While AI promises efficiency gains—Centrica reported a rise in retail profits after focusing on higher-margin services—the human cost is tangible. Call centres, often located in post-industrial towns, have provided steady employment for thousands. Their decline mirrors the broader erosion of middle-skilled jobs, a trend accelerated by automation but rarely acknowledged in the optimistic narratives around AI’s economic benefits.

Datacentres and the new tech arms race

Britain’s role in the global AI infrastructure battle is coming into sharper focus. AMD’s launch of Helios, its first rack-scale AI platform, marks a direct challenge to Nvidia’s dominance in the datacentre market. The system, which AMD claims is the fastest on the market, boasts 72 GPUs and 50% more high-bandwidth memory than Nvidia’s Blackwell-based offerings. At 1.2 metres wide and 44 rack units high, Helios is a physical manifestation of the escalating competition for AI training supremacy.

The timing is critical for the UK. Datacentres already account for 1.5% of Britain’s electricity demand—a figure projected to rise as AI adoption grows. The government’s ambition to make the UK a “science and technology superpower” hinges on its ability to attract and retain such infrastructure. Yet, as Scientific American’s recent analysis of global language loss suggests, technological progress often comes at the expense of cultural and social continuity. The UK’s datacentre boom could exacerbate regional inequalities, concentrating high-skilled jobs in urban hubs while leaving rural areas further behind.

AMD’s pitch is clear: Helios offers better performance at a lower cost. But the real test will be whether UK firms can integrate these systems without deepening their dependence on foreign tech giants. The Bank of England’s recent ban on coal bonds signals a shift toward greener finance, but the environmental impact of AI infrastructure remains a blind spot in Britain’s net-zero strategy.

The ethical tightrope of AI deployment

The UK is not alone in grappling with AI’s societal consequences. In the US, the Cybersecurity and Infrastructure Security Agency (CISA) issued an updated advisory this week warning that Iranian-linked hackers are targeting water and energy providers. The attacks, which exploit vulnerabilities in widely used industrial control systems, underscore the fragility of critical infrastructure in an era of AI-driven cyber threats. For Britain, which has faced its own cybersecurity challenges—including a 2025 ransomware attack on the NHS—the warning serves as a reminder that innovation cannot outpace security.

Meanwhile, OpenAI’s expansion of its ChatGPT Health service to all US users highlights another tension: the personalisation of AI tools. By integrating data from platforms like Apple Health and MyFitnessPal, OpenAI is blurring the line between convenience and surveillance. The UK’s Information Commissioner’s Office (ICO) has yet to weigh in on similar integrations, but the precedent is concerning. As AI becomes more embedded in daily life, the risk of data misuse grows—particularly in sectors like healthcare, where privacy is paramount.

What’s next for UK innovation?

The week’s developments reveal a paradox at the heart of Britain’s tech ambitions. On one hand, the country is positioning itself as a leader in AI regulation and green finance. On the other, its industries are adopting automation at a pace that outstrips policy responses. Centrica’s job cuts, AMD’s datacentre push, and the cybersecurity threats facing energy providers all point to a future where innovation is as much about disruption as it is about progress.

The challenge for the UK will be to ensure that its technological advancements do not come at the expense of its social contract. As AI reshapes industries, policymakers must address the gaps in retraining programmes, regional investment, and cybersecurity. The alternative—a future where efficiency trumps equity—is not one Britain can afford.