AI security cuts UK fleet vehicle thefts by 12 percent

Targa Telematics has reported a double-digit fall in vehicle thefts among UK fleets using its artificial intelligence-based security tools, as demand rises for systems that can intervene before a vehicle is stolen.

The Italy-headquartered connected mobility specialist said fleets equipped with its solutions recorded a 12% reduction in thefts in 2025 compared with 2024. The company attributed the trend to a shift from traditional alarm-based monitoring towards "agentic" artificial intelligence that analyses driver and vehicle behaviour in near real time.

Targa focuses on AIoT, which combines artificial intelligence with networked sensors and devices in vehicles and other assets. The company aims its products at leasing and rental providers, financial institutions and large corporate fleets across Europe and Latin America.

Lower theft incidence

Internal analysis by the Targa Telematics Observatory indicated a broader impact from the new systems. The unit reported a 20% to 30% reduction in theft incidence among comparable fleets that had adopted the tools. The study estimated annual economic benefits of €1.5 million to €2 million for medium and large operators, based on fewer thefts and related losses.

Targa said the results reflected the introduction of enhanced scoring models that assess theft and fraud risk at vehicle level. The company has also redesigned alert workflows for its operations centres. The new processes target fraud and misappropriation in particular, which remain a significant issue for short- and long-term rental providers.

The data points add to wider concerns over vehicle crime in the UK. Fleet and rental businesses have reported rising incidents involving keyless entry systems and organised theft rings in recent years.

Smart Vehicle Protector

A central element in the reported impact is Smart Vehicle Protector, Targa's AI-based product for theft prevention. The product uses behavioural data and contextual information from each vehicle. The technology tracks patterns such as driving behaviour, location histories and use outside agreed contracts.

The system then correlates events and flags complex combinations that can indicate heightened risk. Targa said this gives each vehicle a dedicated layer of monitoring, rather than relying solely on generic alerts across the fleet. Operations centres receive targeted notifications and can intervene earlier in the risk cycle.

The company said this model enables more structured preventive actions and more consistent decision-making by human operators. The aim is to act before a theft takes place rather than only after a vehicle disappears or stops transmitting data.

Shift to agentic AI

Targa has positioned the move as a step beyond earlier generations of telematics and analytics. The firm previously used AI mainly for filtering large volumes of alerts and prioritising those most likely to indicate a genuine incident. That approach focused on reducing false positives and streamlining manual checks.

The new agentic AI systems add a further layer. The models run continuous contextual analysis on each connected vehicle, based on data stored in Targa's data lake. The software can propose or trigger specific actions when it detects behaviour that aligns with known theft or fraud scenarios.

Targa said this does not replace staff in operations centres. The company described the role of the systems as augmenting human judgement, particularly in risk evaluation. The intention is a monitoring model that scales across large fleets while retaining human oversight for high-impact decisions.

The move reflects a wider trend in fleet technology, as providers blend telematics hardware with more sophisticated analytics and automation. Suppliers are seeking to differentiate their services as vehicle connectivity becomes standard on many new models.

Customer impact

The reported economic benefits stem from a mix of avoided vehicle losses, lower recovery costs and reduced disruption to business operations. Medium and large fleets tend to carry high-value vehicles and often operate under service-level agreements with corporate clients, which can magnify the cost of theft incidents.

Insurance relationships may also shift as more detailed risk data becomes available. Fleets using predictive systems can potentially demonstrate lower incidence of theft and fraud over time, although any impact on premiums would depend on insurer assessments.

Targa said it continues to refine its models based on new data from theft attempts and fraud cases. The company expects the combination of telematics history, behavioural analytics and agentic decision logic to adjust as criminals change tactics.

"As the criminal landscape is rapidly evolving, our mission is to deliver solutions that prevent theft and provide our customers with a new layer of security, instead of simply reacting to threats. The application of advanced algorithms to the large volumes of data included in our data lake plays a crucial role in the development and training of increasingly efficient Agentic AI models," said Chris Horbowyj, UK Commercial Director, Targa Telematics. "The new approach we have implemented has allowed us to significantly reduce the economic impact of theft, offering our customers next-generation services that combine security and efficiency through continuous innovation."

Targa operates in several European markets including the UK, as well as in Chile. The company said it plans further development of its agentic AI systems as part of a broader strategy for connected mobility and fleet security.