By Dalibor Z. Chvatal, CC BY 3.0, https://commons.wikimedia.org/w/index.php?curid=56921910
A New Model Emerging: From Mass Surveillance to Intelligent Filtering
Across Europe, the debate around AI-driven video surveillance is still too often framed as a rigid conflict between security and freedom. This framing may be politically convenient, but it no longer reflects technological reality. A more nuanced model is already emerging—one that does not rely on mass surveillance, but on intelligent, purpose-driven filtering of information.
In countries such as Norway, AI-assisted systems are being tested and gradually introduced with a clear strategic direction: avoid intrusive identification, focus instead on recognising patterns and anomalies. Rather than tracking individuals, these systems analyse movements, situations, and irregularities. They detect events such as dangerous traffic behaviour, wrong-way driving, or unusual activity in public spaces without immediately identifying the people involved.
These initiatives are often deliberately embedded in experimental or pilot frameworks, linked to research, urban safety, or anti-terrorism contexts. The intention is not to create automated punishment systems, but to strengthen prevention, situational awareness, and response capability. AI becomes a continuous observer that never tires, scanning large volumes of data and drawing attention only to what truly matters.
This represents a fundamental conceptual shift. The role of AI is not to control society, but to reduce complexity and help authorities focus on relevant risks.
Operational Reality: Where AI Already Works
The most advanced applications of AI-supported surveillance in Europe are not found in broad, generalised policing, but in highly specific and practical domains such as traffic and infrastructure safety.
Norway illustrates this clearly. In road tunnels and transport systems, AI-based solutions are already operational and deliver measurable benefits. They identify stopped vehicles, detect drivers moving in the wrong direction, recognise early signs of fire or smoke, and even flag the presence of pedestrians in restricted areas. These systems operate continuously and reliably, even under difficult conditions such as low visibility or extreme weather.
What makes them effective is not automation of authority, but optimisation of attention. Instead of requiring human operators to monitor endless video streams, the system isolates relevant moments and brings them forward. Human decision-makers remain central, but they are no longer overwhelmed by volume.
Similar developments can be observed elsewhere in Europe. France has experimented with algorithmic video analysis under tightly controlled legal conditions for major events. The Netherlands has introduced intelligent camera systems capable of detecting risky driving behaviour, such as the use of mobile phones behind the wheel. Germany and other countries are exploring anomaly detection in transport hubs and public infrastructure.
The direction is consistent. AI is not replacing human judgement. It is supporting it by making complex environments more readable.
The Legal Reality: Why Progress Is Uneven
Despite these developments, the adoption of AI in surveillance remains uneven across Europe. The reason is not technological limitation, but legal uncertainty.
European frameworks such as the GDPR and the AI Act do not prohibit AI-assisted surveillance. However, they impose strict requirements regarding purpose limitation, proportionality, human oversight, and data minimisation. These principles are not barriers in themselves, but they require clear legal structures at national level.
Countries that are advancing more quickly have understood this. France and the Netherlands have not bypassed European law; they have worked within it by defining precise use cases. AI deployment is limited in scope, often temporary, and always tied to clearly defined objectives. This makes it possible to align technological innovation with legal compliance.
Austria, by contrast, remains more cautious. Existing legal frameworks, including police law and constitutional data protection provisions, were developed before AI-driven analysis became a practical reality. As a result, authorities face uncertainty regarding the permissible scope of automated analysis, the interpretation of proportionality in continuous monitoring, and the legal handling of AI-supported findings.
This hesitation is understandable, but it also reveals a deeper issue. The challenge is not a lack of capability, but a lack of conceptual clarity.
Europe is not blocked by technology; Europe is blocked by legal hesitation and conceptual confusion.
AI as Assistant, Not Authority
At the centre of the debate lies a crucial distinction: the role of AI in decision-making.
A sustainable and legally robust model is already visible. AI systems can analyse vast amounts of visual data, pre-select relevant situations, and indicate potential risks. They can identify patterns that would otherwise remain unnoticed. However, they do not replace human authority. The final evaluation, the legal qualification of a situation, and any resulting decision remain entirely in human hands.
This separation is not a limitation; it is a strength. It ensures accountability, preserves legal certainty, and avoids the pitfalls of automated decision-making. AI becomes a tool that enhances perception and supports judgement, rather than substituting it.
In this sense, AI does not create a new layer of control. It refines existing processes. It helps authorities navigate complexity, identify relevant events more quickly, and use their resources more effectively.
A Practical Framework for Europe
If Europe wants to move forward in a balanced and responsible way, the path is not deregulation but clarification. A clear and transparent framework can enable the use of AI in public security without undermining fundamental rights.
Such a framework would allow automated video analysis for traffic safety and public protection under well-defined conditions. Biometric identification would be excluded, ensuring that individuals are not tracked or identified without cause. Human decision-making would remain mandatory at every stage where legal consequences arise. Data collection and storage would be limited to what is strictly necessary, with strong safeguards to prevent misuse.
This approach is neither radical nor speculative. It reflects the direction already taken in several European countries and aligns with the core principles of European law.
Freedom, Security, and Responsibility
The European reluctance toward AI-assisted surveillance is often justified as a defence of freedom. This argument carries weight, but it also requires careful examination.
Avoiding the use of available technology does not automatically enhance freedom. In some cases, it may simply shift risk onto those who rely on effective public security. Traffic accidents, dangerous behaviour on the roads, and preventable incidents in public spaces are not theoretical concerns. They have real consequences, often for individuals who had no influence over the situation.
AI offers the possibility to recognise risks earlier, to respond more quickly, and to prevent harm more effectively. When used with clear limits and human oversight, it does not expand indiscriminate control. It introduces precision where previously there was only fragmentation.
A society that refuses to use intelligent tools in the name of freedom may, in reality, be leaving its most vulnerable members unprotected.
Conclusion: From Fear to Function
The future of AI in European security will not be shaped by extremes. It will not become a system of total surveillance, nor should it remain paralysed by uncertainty.
The real opportunity lies in a pragmatic understanding of what AI can and should do. It can organise vast streams of information, draw attention to relevant risks, and allow human authorities to act with greater clarity and efficiency. It does not need to replace human judgement to be transformative.
Countries such as Norway demonstrate that it is possible to combine technological capability with respect for rights and transparency. They show that the balance between security and freedom is not fixed, but can evolve with the right framework.
The question for Europe is no longer whether AI should play a role in public security. The question is whether Europe is ready to use it with intelligence, discipline, and purpose.
Foto: By Dalibor Z. Chvatal, CC BY 3.0 Wikimedia
About the Author
Discover more from THE KARDINAL
Subscribe to get the latest posts sent to your email.
