British Gas took 15 months to issue a final bill and refund more than £1,500 to a customer, despite a ruling from the energy ombudsman nearly a year prior. Beth Kojder moved out of her one-bedroom flat in south-east London in October 2024 and subsequently filed a complaint with the ombudsman when British Gas failed to provide a final bill or refund her credit balance.

The energy ombudsman ruled in Kojder's favor in February 2025, instructing British Gas to fulfill her request. However, the ombudsman lacks legal authority to enforce its decisions. Kojder received an offer for her refund this week, shortly before her case was scheduled to be heard in a small claims court.

British Gas stated that it was "implementing the ombudsman's remedy" for Kojder and apologized for the delay. "We are very sorry for how long it has taken," the company said in a statement.

Kojder described the process as "relentless," "tiring," and "completely draining" in an interview with the BBC. She provided meter readings to British Gas when she moved out of her flat to facilitate the creation of a final bill.

The delay highlights ongoing concerns about customer service and billing accuracy within the energy sector. While energy companies are increasingly leveraging AI and machine learning to streamline operations, including billing and customer support, cases like Kojder's demonstrate the limitations of these technologies when not properly implemented or overseen. AI-powered billing systems, for example, are designed to automate the calculation of energy consumption and generate accurate bills. However, these systems rely on accurate data input and robust algorithms. Errors in meter readings, incorrect tariff information, or flaws in the algorithms can lead to inaccurate bills and delayed refunds.

The use of AI in customer service also raises questions about accountability and transparency. Chatbots and virtual assistants are often the first point of contact for customers, but they may not be equipped to handle complex issues or provide personalized support. This can lead to frustration and delays, particularly for vulnerable customers.

The latest developments in AI for the energy sector focus on improving the accuracy and reliability of these systems. Companies are investing in advanced machine learning models that can detect and correct errors in billing data, predict energy consumption patterns, and personalize customer interactions. However, these advancements require careful attention to data privacy, algorithmic bias, and human oversight to ensure fair and equitable outcomes for all customers.