A look back at the numeum social morning on 13 May, hosted by Gwenaëlle Artur and Maud Lambert.

✅ What are the use cases?
✅ What are the risks?
✅ What legal tools are available to regulate the use of AI?
✅ How should you manage information/consultation with the works council?
✅ What clauses should you include in your charters, customer contracts and supplier contracts?

Artificial intelligence (AI) is increasingly being used in HR practices: CV sorting, interview analysis, performance evaluation, etc. These uses promise efficiency and time savings, but raise important legal issues.

🔍 What are the use cases in HR?

• Recruitment: automated sorting of applications, video interviews analysed by AI.
• Talent management: skills assessment, identification of training needs.
• Performance monitoring: analysis of activity indicators or internal communications.
• Absence or turnover management: predictive algorithms.

⚠️ What are the risks?

• Privacy violations (sensitive data, excessive surveillance).
• Algorithmic discrimination (bias related to gender, age, origin, etc.).
• Lack of transparency and explainability.
• Automated decisions without effective human control.
• Employer liability in the event of a dispute.

⚖️ What legal tools are available to regulate AI in HR?

• General Data Protection Regulation (GDPR): regulates the collection, processing and automated profiling of data.
• European AI Act: coming soon, it will classify HR AI as high risk, requiring audits, documentation, transparency and human supervision.
• Labour Code: protection against discrimination, respect for individual freedoms, consultation with the CSE.
• Non-discrimination law: principle of equal treatment.

🗣️ CSE: what information/consultation?

The use of an AI tool that impacts the organisation, working conditions or individual assessments requires:

• Information and consultation with the CSE prior to deployment.
• Communication on the purposes, data processed and consequences for employees.

📝 Clauses to include in your legal documents

• HR charter or AI charter: principles of ethical use, human supervision, employee information.
• Supplier contracts: clauses on GDPR compliance, algorithmic transparency, possible audits, liability in the event of bias.
• Customer contracts (IT services companies, AI publishers): guarantees on data quality, commitments on non-discrimination, right of access and explanation of processing.
• Internal policies: framework for automated processing, procedure in the event of a dispute over an AI decision.