How the EU AI Act Will Reshape the People Counting Market

The European Union’s Artificial Intelligence Act (AI Act), a landmark piece of legislation, is set to fundamentally reshape the technological landscape across various industries. While much of the focus has been on generative AI or high-profile public security applications, one quieter but impactful shift is brewing in the realm of infrastructure management: the evolution of people counting solutions.

For years, the people counting market has been dominated by camera-based devices that leverage embedded AI for superior accuracy in detecting and tracking foot traffic. However, the AI Act’s stringent requirements for digital trust and fundamental rights are poised to disrupt this status quo, creating a massive, and financially compelling, market transition toward non-camera-based alternatives. As detailed in the article Digital Trust in IoT and Smart Buildings, trust is no longer a “soft issue” but a strategic imperative, a new currency in the digital world. This imperative makes inherently private sensor technology an industry prerequisite.

The core of the issue lies in the AI Act’s risk-based framework, which classifies AI systems based on their potential to harm fundamental rights. Many AI-driven camera-based people counters will fall into the ‘High-Risk‘ category, particularly if their underlying system is capable of:

1. Remote Biometric Identification: Even if a system is intended only for counting, its capacity to capture and process personal data (like facial features or gait) brings it under intense scrutiny. The Act includes absolute prohibitions on the untargeted scraping of facial images from the internet or CCTV footage, making the use of raw video feeds for training and operation extremely precarious.
2. Strict Conformity Assessment: High-Risk systems must undergo a rigorous conformity assessment (a mandatory testing and certification process) and adhere to a host of obligations, including implementing a quality management system, maintaining extensive documentation, and ensuring data governance.

Essentially, the core concern is straightforward: AI’s capacity for self-evolution means it could become “unpredictable” and even “uncontrollable” in the future.

The Financial Burden of Certification

The cost of complying with the ‘High-Risk’ classification is a critical factor driving the market switch. Compliance is not a one-time setup; it requires continuous auditing, maintenance, and expert personnel.

• Projected Costs: Early impact assessments estimate that an EU-based Small and Medium Enterprise (SME) that deploys a single high-risk AI system could incur total compliance costs of up to €400,000 (4).
• Price Hike: For a competitive people counting device, adding a portion of this multi-hundred-thousand-euro compliance and certification cost to the product price is untenable. The price of each compliant camera-based device will need to rise considerably, rendering them uncompetitive against simpler, non-camera alternatives.