The EU AI Act's August 2026 Hiring Deadline: What HR Teams Must Do Now
June 15, 2026
Seven weeks. That's how long HR teams have before the EU AI Act's most consequential employment deadline hits on August 2, 2026. If your organization uses AI to screen CVs, rank candidates, or score interviews -- and you employ or hire anyone working in the EU -- you're in scope. And the penalties for getting it wrong are not abstract: up to 35 million euros or 7% of global annual turnover for prohibited uses.
Most HR teams know AI regulation is coming. Far fewer have looked closely at what it actually requires -- or realized how much of their current stack may already be out of compliance.
Here's the practical breakdown.
What the EU AI Act Classifies as High-Risk in Hiring
The Act explicitly designates AI used in recruitment, screening, candidate ranking, interview analysis, skills testing, and performance prediction as "high-risk" under Annex III. That covers a wide swath of modern HR tech: ATS systems with automated shortlisting, one-way video interview platforms with AI scoring, and resume screening tools powered by machine learning.
The scope is also broader than many non-EU employers expect. The Act applies to any organization whose AI outputs are used in the EU -- including a US company hiring a remote employee in Germany. If the AI is making or informing decisions about people working in or from the EU, you're covered.
Source: Greenhouse
What's Already Banned (Since February 2025)
Some things didn't wait for August 2026. As of February 2, 2025, these uses are outright prohibited:
- Emotion recognition in the workplace -- tools that read facial expressions, voice tone, or micro-expressions to infer a candidate's emotional state
- Biometric categorization that infers protected traits like race, political views, or sexual orientation
- Social scoring of candidates' trustworthiness based on broad behavioral data
- Harmful manipulation that exploits psychological vulnerabilities to alter decision-making
If any vendor in your stack offers "read tone/sentiment in video interviews" or "infer personality from facial cues," that feature needs to be disabled now -- not in August.
What Must Be in Place by August 2, 2026
For high-risk hiring systems, the August deadline requires employers (deployers) to have four things in place.
Transparency with candidates. Every person evaluated by a high-risk AI system must be informed that AI is being used and how it influences decisions. The Greenhouse 2026 Candidate AI Interview Report found that 70% of US candidates who experienced AI evaluation were not clearly told beforehand -- a stark gap between where most hiring teams are today and what the law will require.
Worker notification. A separate duty requires informing workers and their representatives before high-risk AI is deployed at the workplace.
Human oversight. Meaningful human review must occur before any final decision. This means a qualified person -- with the authority and competence to override the AI's output -- reviews AI-assisted evaluations before rejections or shortlist decisions are finalized.
Documentation. You'll need records showing who reviewed AI-assisted decisions, what criteria were considered, and how the AI was used. GDPR Data Protection Impact Assessments (DPIAs) also apply where the AI meaningfully shapes hiring outcomes.
Why Human Oversight Isn't Just a Checkbox
The regulation doesn't require that a human is technically "in the loop." It requires meaningful oversight -- someone with the authority to override and the time to actually review.
This matters because the same research found that 63% of job seekers have now experienced an AI interview, and 51% of those candidates never received any outcome after completing one. That silence isn't just a candidate experience failure -- it's exactly the kind of automated dead-end the AI Act is designed to prevent.
Bias is another layer. The Greenhouse report found candidates report nearly identical rates of perceived age bias -- 36% -- from both AI and human interviewers. The Act doesn't eliminate that problem on its own. Structure has to come first: clear criteria, consistent evaluation, and AI used to support human judgment rather than replace it.
What HR Teams Should Do in the Next Seven Weeks
The compliance work isn't complex, but it requires deliberate action before the deadline.
Audit your AI stack. Map every tool that touches candidate screening, ranking, or evaluation. Tag what's high-risk under Annex III.
Disable prohibited features. Any emotion recognition or biometric categorization capability needs to come off immediately.
Update your process disclosure. Add clear language to job postings, interview invitations, and careers pages explaining where AI is used and what it evaluates.
Build a documented human review step. Define who reviews AI outputs, what authority they have to override, and how decisions are recorded.
Check your vendor's compliance roadmap. Providers of high-risk systems must register in the EU database and hold CE marking by August 2026. If your vendor can't confirm this, that's a risk you now own.
The AI Act's intent isn't to slow down AI adoption in hiring -- it's to ensure speed doesn't come at the cost of transparency and fairness. Teams that do this right will find the compliance requirements align closely with what candidates already expect: clear disclosure, structured evaluation, and a human who makes the final call.