How data governance impacts the effectiveness of artificial intelligence in the automotive and insurance sectors, improving data quality and decision-making

Artificial intelligence is increasingly discussed for the opportunities it offers in terms of efficiency, automation and decision support. In the automotive and insurance sectors, AI is already applied across a wide range of use cases, from risk assessment and claims management to driving behavior analysis and process optimization. Less attention, however, is often paid to what truly determines its effectiveness: data quality. Every AI system relies on this essential foundation and, without a solid data governance framework, even the most advanced models risk delivering unreliable or hard-to-interpret results.

Artificial intelligence does not operate autonomously. It learns from data, processes it and uses it to generate predictions or support decisions. In automotive and insurance contexts, this data typically comes from heterogeneous sources, including connected vehicles, telematics devices, sensors, digital platforms and complex information systems. When this information is incomplete, inconsistent or poorly contextualized, AI systems tend to replicate and amplify these issues. The garbage in, garbage out principle — whereby poor-quality data leads to unreliable outcomes — becomes particularly relevant when algorithm-driven decisions directly affect safety, pricing, risk management and customer experience.

It is within this context that data governance becomes a strategic factor. It is not merely a set of technical rules, but a framework that defines how data is collected, managed, controlled and used over time. Effective data governance ensures consistency across different data sources, traceability of information, clear accountability and alignment between data usage and its intended purpose.

The relationship between data governance and artificial intelligence is therefore deeply interconnected. On the one hand, a structured governance approach is a prerequisite for developing robust, explainable and trustworthy AI models. On the other, the growing adoption of AI in critical domains makes responsible data governance increasingly urgent, as algorithms no longer simply support decisions but often help shape them, with tangible impacts on processes, strategies and business models.

The goal is not to accumulate vast volumes of data, but to transform data into high-quality information. In an increasingly complex digital ecosystem, information quality becomes a key factor in building trust in technological systems and enabling truly data-driven decision-making. Governing data therefore means reducing uncertainty, mitigating risk and increasing the value of analytics, making artificial intelligence an effective and sustainable tool for both the automotive and insurance sectors.

In recent years, the amount of data available on mobility has grown exponentially. Sensors, connected vehicles and digital technologies generate millions of data points every day on how we move on the roads.

But there is a fundamental question that often remains in the background: when do data truly become useful to improve road safety?

The answer lies not only in technology, but in method.

Having large amounts of data does not automatically mean making good decisions. Especially when it comes to road safety, available information affects sensitive areas such as public policy, urban planning, insurance models and even citizens’ perception of risk.

Without a rigorous approach, the risk is to oversimplify complex phenomena or to turn data into partial — if not misleading — narratives. It is precisely in this transition, from data to interpretation, that the quality of decisions is determined.

This is why road safety cannot rely solely on aggregated statistics or superficial readings. It requires structured, validated and contextualised analyses.

The scientific method is what allows raw data to be transformed into evidence, intuition into knowledge and observation into decision-making. It is the step that gives meaning to numbers, avoiding simplistic interpretations or narrative shortcuts. Applying a method means questioning data in the right way, taking into account the context in which observed phenomena occur. It means considering territorial differences, road usage conditions, real driving behaviours and the time dimension in which events take place.

Only through this integrated reading is it possible to truly understand when and why risk increases, moving beyond generalisations and standardised solutions that often prove ineffective.

Today, telematics represents one of the richest and most reliable sources of data on real-world mobility. But precisely for this reason, it also carries a responsibility: to use those data correctly, transparently and in a verifiable way.

When technology and scientific research work together, data are subjected to models, controls and criteria that ensure their reliability. It is within this balance between technological innovation and methodological rigour that trust is built.

A rigorous approach makes it possible to shift the focus from simply reacting to accidents to preventing risk. This is a fundamental change in perspective, enabling action before damage occurs, rather than merely measuring it after the fact.

Road safety, in fact, is not just a matter of speed or rules, but of context. It varies according to locations, times of day, traffic conditions and driving habits. Understanding these dynamics means being able to design more targeted, more effective and also fairer interventions, because they are calibrated to the real conditions in which risk emerges.

Along this path, building a new data culture becomes essential. A culture that recognises data as a tool, not an end in itself; that technology must serve people; and that the best decisions are born from solid and responsible analysis.

When data are treated with method, they become a collective resource. When method guides technology, road safety stops being an emergency issue and becomes a structural policy.

Because the transition from data to decisions is never automatic: it is a choice of responsibility.

Cities are changing their skin. Today, talking about Smart Cities no longer means referring only to advanced technologies or experimental projects, but rather describing a new urban model in which data, infrastructure, and public services are integrated to tangibly improve citizens’ quality of life. Just observing how we move every day is enough to understand how central mobility has become to the functioning of cities. In this context, mobility represents one of the most important and transformative domains.

Every movement within a city generates valuable information. Traffic flows, average speeds, travel times, accidents, and the use of public transport and shared services all contribute to building a dynamic map of how urban space is experienced on a daily basis. Sensors, connected vehicles, smartphones, and digital systems feed an increasingly rich information ecosystem, enabling public administrations to observe the city in real time and intervene in a more targeted and effective way.

According to leading international analyses on Smart City development, including the IMD Smart City Index and the EY Smart City Index, the ability to collect and interpret urban data is now one of the key factors in improving the efficiency of public services and overall quality of life. Data becomes a true intangible infrastructure, comparable to physical networks, on which more effective urban policies can be built.

Smart mobility also has a direct impact on safety, sustainability, and social costs. Traffic congestion, road accidents, and pollution represent costs that are often invisible but highly significant for both citizens and public administrations. Conversely, a data-driven approach to mobility management makes it possible to reduce road accidents, cut emissions, and optimize the use of resources, contributing to more livable and resilient cities.

In this scenario, the concept of urban mobility has expanded to include electric micromobility, connected vehicles, intelligent public transport, and sharing services. Digital integration among these different modes of transport supports more flexible and multimodal systems, reducing dependence on private cars and encouraging more sustainable mobility choices.

Smart Cities are also called upon to respond to new challenges related to urban risk. Extreme weather events, infrastructure failures, or emergency situations can disrupt the continuity of essential services. The analysis of mobility and urban data helps identify critical areas, support prevention strategies, and strengthen urban resilience, improving cities’ ability to adapt to complex and evolving scenarios.

In 2025, the world’s most advanced cities demonstrate that the intelligent use of data is no longer an option, but a necessary condition for sustainable urban development. For public administrations, mobility operators, and sectors such as insurance, the Smart City represents an ecosystem in which collaboration, technology, and data converge to create shared value and address future challenges with greater awareness.

________________________________________________________________________________________________________________

Sources
• IMD Smart City Index 2025 – IMD World Competitiveness Center
• EY Smart City Index 2025 – EY Italy
• Intesa Sanpaolo Innovation Center – Smart Cities and urban innovation
• Capgemini Research Institute – Smart Cities Trends 2025