If AI helps you screen, rank or select candidates, the European Union now treats it as high-risk. The deadline has moved, but the direction has not. Here is what changed in 2026 and what employers should ask their recruitment partners now.
AI in recruitment is a high-risk use
The EU AI Act lists AI systems used in employment, including recruitment and hiring decisions, among its Annex III high-risk categories. High-risk systems carry the Act's strictest obligations. For organisations that deploy them, these include ensuring human oversight, monitoring how the system performs, keeping logs, and informing the people affected.
The new timeline: 2 December 2027
The high-risk obligations for Annex III systems were originally due on 2 August 2026. The EU's Digital Omnibus changed that:
- A provisional political agreement was reached on 6 May 2026.
- The regulation was signed on 8 July 2026 and entered into force on 27 July 2026.
- High-risk obligations for Annex III systems, including recruitment AI, now apply from 2 December 2027.
Not everything moved. The Act's Article 50 transparency obligations largely stayed on their original schedule from August 2026.
Breathing room, not a reprieve
Sixteen extra months is time to prepare, not a reason to wait. Employers should map where AI already touches their hiring, including tools used by their recruitment partners.
Five questions to ask any partner using AI
- Where exactly does AI touch the screening? Sourcing, ranking, testing or final selection?
- Who reviews an AI ranking before a candidate is rejected? Human oversight should be real, not a rubber stamp.
- Can you show the audit trail behind a hiring decision? If a regulator or candidate asks, you need records.
- Are candidates told when AI is involved? Transparency builds trust. Gartner found only 26% of candidates trust AI to evaluate them fairly.
- How do you check the model for bias across nationality, gender and age?
What counts as recruitment AI
Many employers already use AI without calling it that. Under a broad reading, the following could involve AI in hiring:
- Software that parses CVs and scores candidates against a job description.
- Tools that rank applicants or filter out candidates automatically.
- Video interview platforms that analyse answers, speech or expressions.
- Chatbots that screen applicants with set questions.
- Matching engines that recommend candidates from a database.
If any of these influence who is shortlisted or rejected, it is worth reviewing them against the high-risk requirements.
Responsibilities for employers and partners
The AI Act distinguishes between providers, who develop AI systems, and deployers, who use them. An employer using a recruitment platform may be a deployer, while the platform vendor is a provider. When you work with a recruitment agency that uses AI, responsibilities can be shared or unclear. Clarify in writing who does what: who monitors the system, who keeps logs, and who informs candidates.
A practical preparation plan
- Inventory: list every tool used in your hiring process, including tools used by agencies on your behalf.
- Classify: identify which tools influence candidate selection.
- Document: record how each tool works, what data it uses and who reviews its output.
- Test: review outcomes for unexpected patterns across groups of candidates.
- Train: make sure recruiters understand when to override an AI recommendation.
Why candidates care
Trust matters in hiring. When candidates believe a process is fair, they are more likely to accept offers and stay. Being open about how AI is used, and showing that a person reviews the result, is good compliance and good employer branding at the same time.
How Mori uses AI
At Mori, AI speeds up the process. It does not replace judgment. AI turns structured job orders into search criteria and helps match and rank candidates, with results tracked in one system. People make the hiring decisions, and every step is recorded in digital worker records with an audit trail.
Want to review how AI is used in your hiring? Talk to us at moriglobal.com/contact.
This article is general information, not legal advice. Consult qualified counsel on your AI Act obligations.
Frequently asked questions
Is AI used in recruitment high-risk under the EU AI Act?
Yes. AI systems used in employment, including recruitment and hiring decisions, fall within the Act's Annex III high-risk categories.
When do the AI Act's high-risk rules apply to recruitment?
After the Digital Omnibus, in force since 27 July 2026, high-risk obligations for Annex III systems apply from 2 December 2027.
Did all AI Act deadlines move?
No. The Article 50 transparency obligations largely remain on their original schedule from August 2026.
Related reading
- The 2027 Compliance Calendar Every International Employer Needs
- Fake Job Candidates: How to Verify Workers When You Hire Across Borders
Sources
- EU AI Act Omnibus agreement: postponed high-risk deadlines (Gibson Dunn, May 2026): gibsondunn.com
- EU AI Act Digital Omnibus now in force (Usercentrics, 2026): usercentrics.com