Sweden and the Nordics
The EU AI Act for HR and people teams
The EU AI Act treats many HR uses of AI, such as recruitment, promotion, task allocation and monitoring of workers, as high-risk. Since February 2025, AI literacy duties and bans such as emotion recognition at work apply. Employers in Sweden should map their HR AI use, classify each case and set up human oversight and logging now.
Book a callMeet us in Stockholm or online. 30 minutes, no sales pitch.
01
Map and classify
List every AI system touching employees and mark which fall under high-risk (Annex III, employment).
02
AI literacy
Article 4 requires staff using AI to have sufficient AI literacy. HR often owns this programme.
03
Oversight and logs
Deployers of high-risk systems must ensure human oversight, keep logs and inform workers and their representatives.
Common questions
- When do the high-risk rules apply?
- They were scheduled for 2 August 2026, and the EU has proposed postponing parts of them. Check the current timeline, but start the inventory now.
- Is emotion recognition at work allowed?
- No, except for medical or safety reasons. The ban has applied since 2 February 2025.
- Do we need to inform the union?
- Deployers must inform workers' representatives before using high-risk AI at work, which fits alongside MBL consultation in Sweden.