--- title: "Article: Automated vehicles and fraud: New technologies, data gaps and new risks" canonical: "/article-automated-vehicles-and-fraud-new-technologies-data-gaps-and-new-risks" published: 2026-11-13 --- # Article: Automated vehicles and fraud: New technologies, data gaps and new risks ## Automated vehicles and fraud: New technologies, data gaps and new risks #### Introduction The arrival of AVs on the UK's roads has the potential to transform transport. Advocates point to safer roads, fewer collisions and greater mobility. Much of the public debate has therefore focused on the benefits AVs may bring. For insurers and fraud professionals, however, a different set of questions arise, one of which is: will AVs reduce fraud, or simply change how it is committed? The answer is likely to be the latter. In particular, new types of fraud are likely to emerge around the gaps in the data that insurers have access to. Fraud has a long history of adapting to technological change. The insurance industry has already contended with organised crash-for-cash operations, manipulated telematics data, synthetic identities and AI-generated documentation. AVs are unlikely to eliminate these risks. Instead, they may create new opportunities for fraudsters to exploit uncertainty, complexity and, crucially, the areas where the available data does not provide a complete picture of what happened. “What the vehicle does not record may become just as important as what it does.” #### AV claims One of the most significant changes AV deployment will bring is the way liability is investigated. In conventional motor claims, disputes typically centre on the actions of human drivers. With AVs, questions may instead focus on whether the vehicle was operating in automated mode, what the automated driving system detected, whether human intervention occurred or should have occurred, and whether a software or system failure contributed to the incident. In principle, vehicle-generated data should help answer those questions. In practice, however, what the vehicle does not record may become just as important as what it does. Not every piece of potentially relevant information will necessarily be collected. Data may be recorded for different purposes, for different periods and to different levels of detail. Some information may be overwritten or unavailable by the time a claim is investigated. Different manufacturers and systems may also capture different types of information. As a result, there may be circumstances in which an insurer has a highly detailed digital record of part of an incident but little or no evidence about another, potentially critical, part. Those gaps may provide the foundations for new forms of fraud. #### Fraud uses gaps Fraudsters are likely to learn quickly what information AV systems capture and share, and, more importantly, what they do not. Claims could increasingly be constructed around events, actions or circumstances that cannot readily be verified from the available vehicle data. Rather than simply fabricating evidence, fraudsters may seek to exploit the absence of evidence, advancing accounts that are carefully designed to fall within known gaps in data collection. AVs generate enormous volumes of information through cameras, sensors, radar, lidar systems and onboard computers. In theory, that data could provide insurers with an objective record of events before, during and after an incident. It therefore has the potential to become one of the most powerful anti-fraud tools the industry has ever possessed. However, more data does not automatically mean complete data, nor does it automatically result in better fraud prevention outcomes. Access to vehicle data remains a significant issue, a subject we explore in detail in our final podcast;for part 1click hereand for part 2click here.Even where relevant information has been recorded, insurers will need to be able to obtain it. Much of the information required to investigate a claim may be controlled by manufacturers or technology providers, meaning clear legal frameworks and workable data-sharing arrangements will be essential. “Fraudsters thrive in environments where evidence is incomplete or difficult to obtain.” This creates two distinct risks. The first is that relevant data exists but insurers cannot obtain it quickly or at all. The second, potentially more significant from a fraud perspective, is that the information needed to test a particular allegation was never collected in the first place. For example, we could see an increased number of claims involving 'passengers' who were not actually present in the vehicle, and insurers will have difficulty defending these claims either because the data recorded does not include passenger information, or that data is not shared with insurers. Fraudsters thrive in environments where evidence is incomplete or difficult to obtain. As AV claims develop, sophisticated fraud may therefore migrate towards precisely those aspects of an incident that fall outside the available data. The industry's understanding of these gaps will need to develop just as quickly as the fraudsters' understanding of them. Insurers need to be especially wary of malicious actors exploiting their hesitancy in incurring extra costs for obtaining data in low-value claims involving AVs. It may not seem reasonable to contest individual claims, but at higher volumes, these will become very costly very quickly. #### Data reliability Even where data is available, interpreting it presents further challenges. Automated driving systems generate highly technical information that may require specialist expertise to understand. Competing interpretations are likely to emerge, particularly in contested claims. Future fraud investigators may spend less time establishing what happened in a conventional sense and more time demonstrating why an alternative explanation advanced by a claimant is inconsistent with the available digital evidence. Inevitably, those investigations will add cost. Data integrity itself may also become a key battleground. If vehicle-generated information becomes primary evidence in liability disputes, there will inevitably be attempts to manipulate, falsify or challenge it. But outright manipulation is only one risk. It may be easier for a fraudster to exploit an ambiguity or omission in the data than to alter the data itself. Fraud may consequently shift away from challenging what the vehicle recorded and towards constructing claims around what it cannot establish. Insurers may increasingly need reliable data not only about the accident but about the condition and functionality of individual components before and after it. Where that information is unavailable, opportunities may arise for exaggerated repair invoices, unnecessary replacement of functioning equipment, questionable recalibration work and inflated associated costs. Establishing whether a sensor genuinely needed replacing may become as important as establishing how the accident occurred. “More data does not automatically mean complete data.” #### Cybersecurity and AI concerns Cyber-related fraud risks are also likely to grow. Connected vehicles increase the number of potential attack points for malicious actors. Allegations of hacking or system interference may become a feature of disputed claims, whether or not any genuine cyber incident occurred. Where the available logs or security data cannot conclusively establish what happened, fraudsters may attempt to use that uncertainty to introduce alternative explanations that are expensive and technically difficult to disprove. We explore cybersecurity and hacking concerns in greater detail in a separate podcast. AI adds a further dimension. Insurers are already exploring how AI can assist with processing large volumes of data, identifying patterns, detecting anomalies and supporting fraud triage at scale. But the same technology is available to fraudsters. AI tools can be used to create convincing documentation, generate credible but ultimately false claims narratives and support sophisticated identity fraud. AI-based fraud detection is only as effective as the information available to it. If there are systematic gaps in AV data collection, those gaps may become blind spots not only for human investigators but also for automated fraud detection systems. Fraudsters who identify those blind spots may be able to adapt their claims accordingly. “AV data could become one of the most powerful anti-fraud tools the industry has ever possessed.” #### Conclusion The long-term effect of AVs is likely to be a reduction in accident frequency, particularly where automation successfully reduces human error. Fewer accidents should mean fewer opportunities for certain established types of fraud. However, reduced claim volumes do not necessarily translate into reduced fraud risk. Fraud may instead become more technologically sophisticated, more expensive to investigate and increasingly concentrated in areas where the available data cannot provide a definitive answer. AVs may change who is behind the wheel, or even the presence of a wheel, but they are unlikely to change a fundamental truth of insurance fraud: wherever uncertainty exists, fraudsters will look for an opportunity. In an AV environment, some of the most important uncertainty may be created not by a lack of data generally, but by specific gaps in what is collected, retained and accessible. The key challenge for insurers will therefore be preparedness. Access to data, technical expertise, cybersecurity capability and robust fraud investigation processes will all become increasingly important as AVs are deployed across the UK. But insurers will also need to understand the limits of the available evidence. That means asking not only what data do we have, but what data is missing, why is it missing and how could that absence be exploited? Identifying those gaps early may be critical to anticipating the next generation of fraud typologies rather than discovering them only after fraudsters have begun to exploit them. “Fraud may migrate towards precisely those aspects of an incident that fall outside the available data.” Claire Newcombe Partner DAC Beachcroft, CSG Rob Fallows Head of Intelligence DAC Beachcroft, CSG **Next page >>** **Careers** **Complaints / Our Policies** **Accessibility** **Legal & Regulatory** **Emergencies** **Modern Slavery Act Statement** 2025 DAC Beachcroft LLP. All rights reserved.