Deleuze replied: From a legal standpoint, the first risk is unlawful processing. IRIS cannot simply collect heart-rate data because it improves fatigue detection. The controller must show that the processing is lawful, necessary, proportionate, and limited to the stated purpose of driver drowsiness detection. If explicit consent is used, it must be genuinely free and informed, which may be difficult in employment, fleet, insurance, or public transport settings where drivers may feel they have no real choice. The second legal risk is purpose creep. Heart-rate data collected for fatigue detection could be misused to infer stress, emotional state, health conditions, fitness for work, productivity, or insurance risk. That would go beyond the original road-safety purpose and could breach GDPR’s purpose limitation and data minimisation principles. The safer approach is to define the purpose narrowly as real-time fatigue detection and prohibit secondary uses unless separately justified.

Deleuze replied: For sure. Heart rate is not a perfect indicator of drowsiness. It can vary because of exercise, caffeine, anxiety, illness, medication, disability, pregnancy, temperature, or cardiovascular conditions. If IRIS treats elevated or reduced heart rate as fatigue without context, it may produce false positives or false negatives. Ethically, this could unfairly burden some drivers or fail to protect others.

Deleuze replied: Definitely. Heart-rate data is sensitive and could be harmful if leaked or sold and misused. IRIS would need encryption, access controls, audit logs, separation of identifiers, secure deletion, and controls preventing access by employers, insurers, fleet managers, or third parties unless strictly necessary. GDPR requires security appropriate to the risk, including measures such as pseudonymisation and encryption where appropriate. Plus, drivers should be told what heart-rate data is collected, how it is used, whether it is stored, who can access it, how long it is retained, and what consequences may follow from a fatigue alert. They should also have a way to challenge inaccurate or unfair alerts, especially if the data is used in employment, insurance, disciplinary, or safety decisions.