Recrutement et entretiens assistés par IA

Ce qu'elle couvre

Ce programme de niveau praticien permet aux recruteurs et aux responsables du recrutement d'intégrer des outils d'IA à chaque étape du processus de recrutement — du traitement des CV et de la notation des candidats à la génération de grilles d'entretien structuré et à la synthèse des notes post-entretien. Les participants apprennent à identifier et à atténuer les biais algorithmiques, à appliquer les exigences du Règlement IA européen relatives aux systèmes à haut risque, et à construire des processus auditables et conformes. Le format combine des ateliers en direct basés sur des cas réels et des sessions pratiques avec les outils du marché.

À l'issue, vous saurez

• Configure and critically evaluate an AI CV-screening workflow, including defining scoring criteria and fallback rules
• Generate a structured interview guide using an LLM, then audit it for legally protected-characteristic bias before deployment
• Apply the EU AI Act Article 10–14 obligations to a recruitment AI system and produce a basic conformity checklist
• Detect and document algorithmic bias in a candidate-ranking output using provided audit templates
• Design a human-in-the-loop approval gate that satisfies both operational efficiency and explainability requirements

Sujets abordés

• AI-powered CV parsing, ranking, and candidate shortlisting
• Structured interview generation and AI-assisted question banks
• Post-interview note synthesis and summary tools
• Algorithmic bias detection and fairness auditing in hiring
• EU AI Act high-risk classification and obligations for recruitment systems
• GDPR and data retention rules for candidate data in AI pipelines
• Human-in-the-loop decision governance for hiring decisions
• Vendor evaluation criteria for compliant AI recruiting tools

Modalité

Delivered as a blended programme over two to three weeks: two half-day live virtual workshops plus asynchronous assignments between sessions. An optional in-person capstone day is available for groups of 10 or more. Participants work on a real or anonymised hiring funnel from their own organisation. Approximately 60% of time is hands-on with tools (e.g. ChatGPT/Claude for interview design, HireVue or Ashby demo environments, Excel-based bias audit templates). Printed job aids, a compliance checklist, and a bias-detection playbook are included.

Ce qui fait que ça marche

• Establishing a clear policy document before roll-out that defines which decisions AI may inform versus make autonomously
• Running a bias audit on a historical hiring dataset before deploying any scoring model to establish a baseline
• Training both recruiters and hiring managers together so governance expectations are shared across the full funnel
• Scheduling a 90-day post-deployment review to assess disparate-impact metrics and recalibrate scoring criteria

Erreurs fréquentes

• Treating AI candidate scores as final decisions without a documented human review step, creating EU AI Act non-compliance
• Using a general-purpose LLM to rank CVs without defining structured, job-relevant scoring criteria — amplifying existing hiring biases
• Failing to inform candidates that AI is used in the screening process, violating GDPR transparency requirements
• Buying an AI recruiting tool without checking whether the vendor provides the technical documentation required under the EU AI Act for high-risk systems

Fournisseurs à considérer

• SHRM (Society for Human Resource Management)www.shrm.org/education/educational-programs/ai-in-hr →
• Coursera – AI in Human Resources (DeepLearning.AI)www.coursera.org/learn/ai-in-human-resources →
• DataScientestdatascientest.com/formation-ia-rh →
• LinkedIn Learning – AI in Recruitingwww.linkedin.com/learning/topics/recruiting →

Sources

• EU AI Act – High-Risk AI Systems in Employment (Annex III) →
• OECD – OECD Principles on AI and Hiring →