Loading…
Loading…
Employment and recruitment AI is Annex III high-risk under the EU AI Act. Archiet generates Art. 10–15 documentation, human oversight scaffolding, and production-ready HR code from one formal model — not a compliance PDF that diverges from what you ship.
Not sure of your tier? Classify your HR AI system free, then read the EU AI Act high-risk compliance guide or Annex IV documentation workflow.
High-risk provisions for Annex III systems apply from 2 August 2026. Documentation and oversight cannot be retrofitted after launch.
AI hiring tools are high-risk under Annex III
Archiet classifies recruitment and performance AI as EU AI Act high-risk and generates the Art. 10–15 compliance pack alongside your codebase.
Art. 14 human oversight is an afterthought
Human override queues, recruiter stop controls, and audit trails are modelled in the genome and emitted in generated code — not documented only in a PDF.
Training data and bias evidence is missing
Article 10 data-governance sections and training-dataset documentation templates ship in every high-risk HR bundle.
August 2026 deadline leaves no runway
Start from the HRM ERP template with eu_ai_act_high_risk flags pre-set. Conformity checklist and post-market monitoring plan included.
Template baseline quality score: 89. Includes applicant screening models, human override endpoints, and full EU AI Act artifact pack.
Ship applicant scoring with mandatory human review queues and EU AI Act documentation before the August 2026 high-risk enforcement date.
Extend existing HR systems with performance-risk models that include oversight mechanisms and regulator-ready technical files.
Deliver client HR platforms with compliance artefacts auditors can open — not slide decks that drift from the shipped code.
Yes. Employment and recruitment AI falls under Annex III high-risk categories. Archiet pre-classifies HRM genomes with eu_ai_act_high_risk flags and emits Art. 10–15 documentation alongside the generated codebase.
High-risk provisions for Annex III systems apply from 2 August 2026. Documentation, human oversight, and conformity evidence must exist before market placement — not retrofitted after launch.
Risk tier assessment, high_risk_hr_articles_10_15.md, Annex IV technical documentation, transparency notices (Art. 13), human override API endpoints, conformity checklist (Art. 43), and post-market monitoring plan (Art. 72) — all from the same genome as your HR software.
Yes. Human override queues, recruiter stop controls, and audit trails are modelled in the genome and emitted in generated code — not documented only in a separate PDF.
Yes. Use the free EU AI Act Risk Classifier — three steps, no account required. If you are high-risk, follow the readiness checklist, then generate the full bundle from the HRM template.
All 12 Archiet web stacks including Flask + Next.js, FastAPI, Django, NestJS, Laravel, Rails, Go, Java Spring Boot, and .NET — same HR architecture, your team's preferred stack.