How recruitment agencies can deploy AI without breaching the EU AI Act, NYC Local Law 144, Illinois AIVIA, and a dozen other regimes. Jurisdiction-mapped and operationally specific.
Recruitment is the fastest AI-adopting professional services sector — and the most regulated. From 2 August 2026, the EU AI Act classifies recruitment AI as high-risk, demanding conformity assessments, technical documentation, and human-oversight regimes most agencies have never built.
NYC Local Law 144 has required independent bias audits since 2023. Illinois's AIVIA, in force from January 2026, attaches USD 5,000 per-candidate penalties to non-consensual video-interview AI. The UK ICO has accepted 296 recommendations from AI recruitment vendors and now finds the majority of automated recruitment decisions are unlawful.
This 73-page playbook compiles the published rules, operational patterns, and vendor evidence agency owners need to deploy AI without landing on the wrong side of any of them. It covers the full source-to-shortlist workflow, names the specific behaviours regulators are penalising, and maps obligations across the EU, UK, US states, Quebec, Singapore, and Australia. Reference material for principals and compliance leads.
73 pages · 12 sections
What's inside
- 1.Executive Summary
- 2.How Recruitment Actually Works (Operational Baseline)
- 3.AI Agent Use Cases in Recruitment
- 4.Core Risk Categories
- 5.Jurisdiction-by-Jurisdiction Compliance Analysis
- 6.AI Vendor Risk Analysis
- 7.Do Not Do This: High-Risk Behaviours
- 8.Best Practice Architecture (Safe Implementation Model)
- 9.Practical Implementation Playbook
- 10.Recruiter Behaviour: What Must Change
- 11.Future Outlook
- 12.Legal Notice