High-Risk Classification
Under the EU AI Act (Article 6, Annex III), AI systems used in recruitment and employment contexts are classified as high-risk. This applies to tools that assist in screening, evaluating, or ranking candidates — including AI-powered interview analysis platforms like Hirify. High-risk classification means that both the AI vendor and the deploying organisation (your recruitment team) must ensure:- Transparency to candidates — individuals must be informed when AI systems are used to assess them.
- Explainable outputs — AI-generated assessments must be traceable and understandable, not black-box scores.
- Human oversight before any decision is made — no AI output should directly determine a hiring outcome without a human reviewing it first.
- Logging and auditability — the system must maintain records of AI-generated outputs sufficient to support post-hoc review and regulatory inquiry.
The AI Act’s high-risk provisions for hiring tools apply from August 2026. Hirify is designed to be compliant from day one, so your team is ahead of the deadline rather than scrambling to catch up.
How Hirify Complies
Hirify’s product architecture directly addresses each high-risk requirement:- No opaque scoring — Hirify does not produce unexplained scores or rankings. Every structured data point in a candidate profile is linked to the interview quote or exchange that generated it. You can always see why a field was populated.
- No automated decisions — Hirify surfaces structured information and AI-generated summaries as inputs to recruiter judgment. The platform does not automatically advance, reject, or rank candidates. Every consequential action requires a human to act.
- Full audit trail — all enrichments are timestamped, attributed to the interviewer, and tied to a specific interview session. If a candidate or regulator asks what data was generated and when, you have a complete record.
- Candidate transparency — candidates can request to see all data Hirify holds about them via rgpd@hirify.fr. This right is aligned with both the AI Act’s transparency requirements and GDPR Article 15.