Enabling Trust and Compliance in Driver Monitoring with Lossless Data Anonymization

The automotive industry is at a critical inflection point where regulatory pressure,
technological capability, and user expectations are converging. The introduction of
mandatory driver monitoring under EU General Safety Regulation (Regulation (EU)
2019/2144), combined with the entry into force of the EU AI Act (Regulation (EU) 2024/1689)
has fundamentally reshaped how in-cabin systems must be designed, deployed, and
governed. At the same time, heightened public awareness around surveillance and data
misuse is influencing driver perception and acceptance of these technologies.
The regulatory landscape presents a structural tension. Driver monitoring systems must
continuously analyze driver behaviour to meet safety requirements, yet the same systems
are classified as high-risk AI systems under Annex III of the EU AI Act, triggering obligations
around data governance, transparency, robustness, and human oversight. In parallel, the
GDPR (Regulation (EU) 2016/679) classifies the sensitive data in-cabin systems inherently
process, including facial geometry, gaze patterns, and behavioural signals, as special
category personal data under Article 9, demanding the highest level of protection. This dual
requirement creates a compliance dilemma that cannot be resolved through policy
interpretation alone.