The European Union (EU) recently reached a provisional agreement on amendments to the EU Artificial Intelligence (AI) Act that, if formally adopted, will delay the requirements for “high-risk” AI systems, including those used to make employment decisions, from taking effect on August 2, 2026, until December 2, 2027.
Quick Hits
- The European Union has provisionally agreed to delay the implementation of high-risk AI regulations in employment decisions until December 2, 2027.
- Subject to formal adoption, the agreement would delay the requirements for “high-risk” systems, which include those intended to be used to make decisions affecting employment terms, promotion, termination, task allocation, and monitoring or evaluating workers’ performance.
- Although the revised deadline is expected, employers may still want to prioritize preparing for compliance by the original deadline considering the scope of the EU AI Act’s requirements.
New Deadline Pending
On May 7, 2026, the European Parliament, Council of the European Union, and the European Commission reached a provisional agreement on a proposal, the Digital Omnibus on AI, to amend parts of the EU AI Act. On June 16, 2026, the European Parliament approved the provisional agreement, which now must be formally approved by the Council before it becomes law.
Notably, among the changes, the proposal would delay the application of the AI Act’s requirements for high-risk AI systems, which include those used in employment, until December 2, 2027.
Once the agreement is approved by the Council, the amendments would enter into force three days after publication in the Official Journal of the European Union. Publication is expected before August 2026.
For employers, the key point is that the delay is not a repeal. Employment-related AI remains squarely in the AI Act’s high-risk category when used for recruitment, selection, promotion, termination, task allocation, or performance monitoring. Although the new deadline is expected to become official beforehand, employers may want to continue preparing for the August deadline out of an abundance of caution.
Regulation of High-Risk Systems
The EU AI Act takes a risk-based approach that subjects AI tools to a spectrum of four levels of increasing regulation based on perceived potential risks: (1) “unacceptable risk,” which are banned; (2) “high risk”; (3) “limited risk”; and (4) “minimal risk.”
Under Annex III of the AI Act, high-risk employment systems include AI used to place targeted job advertisements, analyze and filter job applications, and evaluate candidates. They also include AI used to make decisions affecting employment terms, promotion, termination, task allocation based on individual behavior or personal traits, and monitoring or evaluating workers’ performance or behavior.
Overall, the AI Act requires employers to notify workers that a high-risk AI system is used in the workplace, provide human oversight and intervene when necessary, monitor for potential discriminatory impacts, record AI system logs, and ensure compliance with data privacy requirements. Employers using a vendor’s AI system for the employment purposes described are also required to follow the vendor’s instructions, which vendors are legally required to provide.
Next Steps
The provisional agreement and the delay in compliance deadlines may reflect an acknowledgment that more time may be needed to adapt to the new regime. If formalized, this will provide additional time to make the necessary compliance adjustments.
However, the changes merely delay certain deadlines and do not substantially change impending compliance obligations. High-risk workplace AI remains regulated. Employers may wish not to wait and begin preparing now, including reviewing which AI tools they use and how they are used, in order to develop and implement a compliance plan.
