Self-assessment reportSelf-assessment report

3 AI systems assessed for a 40-person recruiting firm (deployer). 1 high-risk (Annex III), 2 limited-risk. 4 obligations binding on August 2, 2026; 2 deferred. 30/60/90-day plan below.

CV screening tool: HIGH-RISK (Annex III, 4: employment). Support chatbot: limited-risk (Art. 50 transparency). Marketing content generator: limited-risk (Art. 50 labelling of AI-generated content).

Chatbot: inform users they are talking to an AI (Art. 50). Generated content: AI labelling (Art. 50). CV screening: deployer duties under the Annex III regime, human oversight and log retention (Arts. 26, 14, 12).

Extended documentation duties tied to harmonised standards still being finalised, and obligations for AI in regulated products with longer transition windows. Dates and articles listed per system in the full report.

30 days: add the AI disclosure to the chatbot, label generated content. 60 days: human-oversight procedure and logging for CV screening, ask your vendor for their Annex IV documentation. 90 days: review supplier contracts and set a re-assessment routine.

The Act keeps phasing in after August 2. Compliance Monitoring (39 €/month) watches the changes and flags the ones that affect these three systems.