Ethical and Compliant Analysis
In the Career Lab, artificial intelligence helps you improve the way you present yourself: it does not judge you. The analysis describes observable communication behaviours — how you structure your speech, the pace at which you speak, how you address the camera — and turns them into practical suggestions that can be acted upon and improved. It does not infer your emotional states, does not assign a score to the person, and never replaces the teacher's judgement.
AI supports the learning journey: it observes communication elements, helps reflection and leaves final interpretation to tutors and teachers.
The guiding principle: describe, do not label
Two rules follow from this principle and apply across the entire platform:
We observe, we do not diagnose
The platform describes observable behaviours and offers points for reflection; it does not make judgements about a person's inner state or provide psychological "diagnoses".
The final judgement remains human
The final interpretation remains with you, together with teachers and tutors. AI prepares data and feedback; people make the decisions.
What we analyse — and what we never do
The analysis always remains focused on observable behaviours and communication skills. We never attribute emotional states or personality traits.
| Mode | What we analyse — observable | What we never do |
|---|---|---|
| Text / CV and answers | Vocabulary range, correctness, coherence and structure of speech, clarity, relevance of answers, use of the STAR method | Labels such as "self-confident" or "motivated"; inferences about personality or psychological traits |
| Voice | Speaking speed, frequency and duration of pauses, volume, intonation variation, fluency, fillers such as "um" | States such as "nervous", "anxious" or "stressed" inferred from the voice |
| Video | Posture, frequency of gestures, gaze direction towards the camera, head orientation, presence within the frame | "Insecure", "uncomfortable", micro-expressions, emotions read from the face |
AI organises useful observations, but feedback remains formative and supervised by tutors and teachers.
Why we do it this way: the European rule
The Career Lab operates in an educational context and supports preparation for selection processes. In this context, European legislation on artificial intelligence — Regulation (EU) 2024/1689, the "AI Act" — prohibits the use of systems that infer a person's emotions from biometric data. This prohibition has applied since 2 February 2025.
There is, however, a clear distinction: inferring an emotion or inner state is prohibited; detecting immediately evident expressions, gestures and movements is permitted. Our platform remains entirely on this second side: it measures how you communicate, not what you feel.
This choice is not only a matter of compliance: it is also more robust from a scientific perspective. European legislation itself recognises the limited reliability of systems that claim to read emotions from the face or voice. Inferring "anxiety" from an expression is fragile and unlawful; measuring speech rhythm is robust and lawful.
The tutor reviews the report, validates it and turns it into a formative conversation with the student.
Human supervision is always present
No feedback generated by AI reaches the student without tutor validation.
The teacher reads the report, may edit it, and approves it before delivery. AI is a tool for shared reflection, not a verdict.
Words such as "confidence" or "engagement": how we handle them
Terms such as confidence or engagement are ambiguous: if presented as "measured" mental states, they would fall within the prohibition. For this reason, we do not assign scores to inner states. We describe observable behaviour — for example, "steady eye contact, stable posture, regular pace" — and leave it to the student, together with teachers and tutors, to draw conclusions.
Regulatory references
- AI Act — Regulation (EU) 2024/1689, Article 5(1)(f) and Recital 44
- European Commission Guidelines on prohibited AI practices, February 2025
- GDPR — Regulation (EU) 2016/679, Article 9, biometric data as a special category of personal data
This page is for information purposes and describes the platform's approach; it does not constitute legal advice. Data processing takes place according to the measures described in the Privacy and Data Security section.