Generative AI on the Forecourt - Navigating Opportunity and Risk in Fuel and Payments

The fuel industry is constantly changing, driven by new technology and consumer behaviour shifts. While electric vehicles mark a fundamental change in energy use, the immediate challenges and opportunities for forecourt operations and payment systems continue to evolve. Generative AI (GenAI) is emerging within this dynamic environment as a technology that can significantly improve efficiency, fraud detection, and customer engagement.

GenAI is already making waves in the fuel industry, particularly in optimising processes. So far, its capabilities include generating draft documents, presentations, and computer code, as well as summarising large documents and emails. While these applications are just the beginning, we anticipate even greater efficiency as the technology matures and integrates more deeply.

This is where Agentic AI comes in, taking GenAI to the next level by enabling intelligent process automation. Instead of just generating content, Agentic AI can understand a problem and determine the best course of action. This capability is perfect for a wide range of back-office operational tasks. For instance, imagine a fraud analyst investigating a suspicious transaction. An Agentic AI could analyse various data points, such as transaction history, customer profiles, and location data, and then autonomously decide whether to flag the transaction for human review, block it, or request further verification from the customer – all based on its assessment of the risk.

GenAI's spread isn't limited to legitimate business uses. Criminals are also harnessing its power to ramp up fraudulent activities. They now use GenAI to create highly realistic fake fuel card applications, often getting the necessary stolen data from illicit online marketplaces. The resulting fraudulent cards can cause substantial financial damage when these counterfeit applications are approved. What's more, these fraudsters are experts in fuel card fraud, enabling them to exploit the cards for maximum value over extended periods, often without immediate detection, which significantly escalates the overall problem.

In response to this growing threat, fraud detection tools increasingly incorporate GenAI capabilities to strengthen their defences. New GenAI-driven features allow for advanced fraud screening and detection, helping identify suspicious activities earlier. GenAI is particularly good at processing sequential data and spotting subtle differences between legitimate and fraudulent patterns, often leading to much earlier identification of fraudulent transactions than previous analytical methods. Additionally, the development of Agentic AI has the potential to change the way we build fraud systems, continuously adapting to new fraud trends, requiring minimal human intervention or feedback, such as insights from fraud management teams during case reviews. This ability to learn and adapt is crucial in an environment where fraud tactics constantly evolve.

While sophisticated technological safeguards are critical for reducing cyber threats, human elements often remain vulnerable. Fraudsters are aware of that and have capitalised on it using GenAI to create very convincing phishing emails. These communications are designed to trick staff into clicking malicious links, opening harmful files, or even directly transferring funds. Soon, GenAI-based email scanners are expected to mitigate this threat. These systems will analyse email content and, by cross-referencing it with past communications from the supposed sender, determine how likely the message is to be authentic. For instance, if there's an increase in phishing attempts impersonating senior leadership, an email scanner could compare the current message against all previous emails from that individual, providing a probability score regarding its legitimacy.

Beyond the immediate concerns of fraud, the gradual shift towards electric vehicles is causing a fundamental change in how forecourts operate. As internal combustion engine vehicles become less common, forecourts adapt to the longer dwell times associated with EV charging. This shift is driving the integration of additional services, transforming forecourts into multi-purpose destinations. GenAI is poised to become a core technology for the strategic deployment and continuous management of modern forecourt ecosystems. Its analytical power will inform key decisions, from optimising product placement in retail areas to determining the ideal type and number of additional facilities needed at various forecourts, depending on foot traffic and geographical location.

The increasing number of customers on-site, especially during busy charging periods, creates new operational challenges. A significant concern is congestion at EV charging points, which can lead to customer frustration and a potential loss of business if all charging infrastructure is occupied. GenAI can continuously monitor site traffic data, enabling more efficient operational management. This includes dynamically adjusting staffing levels to match demand. When integrated with fuel card platforms, this can be further enhanced through real-time, personalised offers delivered directly to drivers. Leveraging vehicle telematics data, the fuel card platform can ascertain a vehicle's current location and anticipate its planned route and remaining fuel or energy levels. This allows for automated route updates with targeted incentives, encouraging drivers to stop at specific sites through tailored fuel or site-specific promotions, thereby improving the overall operational efficiency of the forecourt network.

Fuel card operators themselves stand to benefit significantly from advancements in GenAI. GenAI can provide early warnings of potential customer churn, payment defaults, or shifts towards rival fuel card providers by analysing individual customers' and cards' regular fuelling patterns. This predictive capability allows operators to intervene proactively, for example, by issuing personalised offers designed to foster continued customer loyalty and usage. Furthermore, customer behaviour can be monitored more frequently to ensure each customer receives the most appropriate discount plan for their usage profile. This granular level of customer engagement, informed by GenAI, can contribute to deeper customer loyalty and sustained business relationships within the evolving fuel and payments ecosystem.

Generative AI (GenAI) is already boosting efficiency, and it's set to integrate into nearly every business area in the coming years, fuelling significant growth. This technology's future is both exciting and disruptive as it advances. One certainty is that GenAI will enable businesses to respond faster to evolving trends and dramatically improve back-office operations, leading to greater customer satisfaction and business expansion.