The EU AI Act has reached an important point in 2026, but some of the dates and requirements have changed from what many businesses were preparing for earlier this year. If you use AI in recruitment, here’s what has actually changed, what has not, and what you should be doing with the extra time.
The Headline Change: The EU AI Act Deadline Moved
The original application date for the EU AI Act rules covering high-risk AI systems was 2 August 2026. That has now changed.
Under the AI Omnibus, the rules for high-risk AI systems covered by Annex III will apply from 2 December 2027. Employment is specifically included among the high-risk areas covered by these rules.
So yes, the EU AI Act deadline moved. But the underlying requirements have not simply disappeared. Organisations still need to prepare for requirements around risk management, documentation, data quality, logging, human oversight and accuracy.
What High-Risk AI Still Means for Recruitment
Recruitment and employment remain within the high-risk areas covered by the AI Act high risk framework.
That matters if you use AI to help with tasks such as screening, ranking or evaluating candidates. The delayed date gives organisations more time, but it does not change the direction of the EU AI Act requirements.
And the AI Act is not a replacement for GDPR. Your obligations around candidate data and automated decision-making need to be considered separately.
The Fine Structure Is Worth Getting Right
There has also been confusion around AI Act fines.
Under the regulation, certain violations of the obligations applying to providers and deployers can result in fines of up to €15 million or 3% of worldwide annual turnover, whichever is higher. A separate maximum of €35 million or 7% of worldwide annual turnover applies to prohibited AI practices.
These are different categories, so they should not be presented as though every EU AI Act compliance issue carries the €35 million maximum.
Why the New Deadline Doesn't Mean You Can Wait
A later deadline can make it tempting to put the EU AI Act on the back burner.
That would waste some of the most useful preparation time you have.
Documentation, human oversight and testing are not things you want to build into a recruitment workflow a few weeks before a deadline. The extra time gives you an opportunity to understand how your AI tools actually work and identify gaps before the rules apply.
EU AI Act Compliance Is Not Just About the AI
For recruitment teams, EU AI Act compliance should not be treated as a checklist that starts and ends with the software itself.
Your recruitment process matters too.
How candidates are screened, who reviews recommendations, what information is used and how decisions are recorded can all affect how AI is used in practice. Looking at the complete workflow gives you a much clearer picture than simply asking whether a vendor says its product is compliant.
If you're also reviewing your recruitment tools, it's worth looking beyond compliance claims and checking how the software actually handles candidate data. Our guide on How to Choose AI Recruiting Software Putting Privacy First covers what to look for.
Where EdgeTal Stands
EdgeTal takes a different approach to handling candidate data.
Candidate data used for AI processing stays on the device rather than being sent to a cloud server for AI processing. EdgeTal also uses a private LLM locally on the device for its matching and search functionality.
That can reduce exposure around cloud processing and third-party data transfers. But it does not automatically make a recruitment tool compliant with the EU AI Act.
Requirements around documentation, risk management, human oversight, data quality and other controls still need to be addressed separately. Privacy architecture can support compliance, but it is not a substitute for the full process.
Conclusion
The EU AI Act deadline for high-risk AI systems in areas such as employment has moved to December 2027. That gives recruitment teams more time, not a reason to stop preparing. Use the extra runway to understand your AI tools, document how they are used and build proper human oversight into the recruitment process.
FAQs
What is the EU AI Act?
The EU AI Act is the European Union's legal framework for regulating artificial intelligence based on its level of risk. It introduces specific requirements for high-risk AI systems used in areas including employment.
When is the EU AI Act deadline for high-risk recruitment AI?
The rules for high-risk AI systems covered by Annex III, including relevant employment use cases, will apply from 2 December 2027.
What are the main EU AI Act requirements for recruitment?
The EU AI Act requirements include areas such as risk management, data quality, documentation, logging, human oversight, accuracy and cybersecurity.
Is recruitment AI automatically high-risk?
Not every AI system used in recruitment is automatically high-risk. The classification depends on how the system is used and whether it falls within the high-risk use cases defined by the regulation.
Does GDPR still apply to AI recruitment?
Yes. The EU AI Act and GDPR address different areas and can apply alongside each other. Recruitment organisations still need to consider their obligations when processing candidate data and using automated systems.