EU AI Act
The EU AI Act is the European Union regulation on artificial intelligence, which sets rules for AI systems according to the level of risk they present to people's safety and fundamental rights.
Definition
The EU AI Act is a European regulation that governs how artificial intelligence systems are developed and used in the Union. Its central idea is a risk-based approach: the higher the potential harm of a use, the stricter the rules. Some practices are prohibited, certain uses are classed as high-risk and carry detailed obligations, some must meet transparency requirements, and most uses face few specific rules.
Employment is one area the regulation treats with particular care. AI used in recruitment or to evaluate people at work is generally considered high-risk. For such systems the main themes are risk management, good quality and representative data, documentation, transparency towards users, human oversight and accuracy and security. Where people interact directly with an AI system, they should be informed that it is an AI.
For organisations, the practical step is to list the AI tools they use, work out which category each falls into, and ask providers what documentation and safeguards they offer. Keeping a person in charge of decisions helps in any case. The timing of obligations is phased, and details change, so do not rely on this page alone and check the current requirements with a legal advisor.
Key points
- It regulates AI in the EU with a risk-based approach.
- Some uses are prohibited, others are high-risk with detailed obligations, others only need transparency.
- AI used in recruitment and evaluation of workers is generally treated as high-risk.
- Main themes: risk management, data quality, transparency, human oversight and security.
- Obligations are phased in; check current requirements with a legal advisor.
Example
An HR department lists the AI tools it uses: a CV sorting tool, an interview assistant and a chatbot for employees. For the first two, which influence hiring, it asks the providers about their documentation and human oversight features, and it checks with legal counsel how its obligations as a user apply.
Frequently asked questions
Who must comply with the EU AI Act?
Providers who develop AI systems and organisations that deploy them have different obligations. Which role you hold, and which rules apply, depends on your situation.
Is the EU AI Act the same as the GDPR?
No. The GDPR protects personal data, while the AI Act regulates AI systems according to their risk. A tool that uses AI on personal data may need to meet both.
Does it ban AI in recruitment?
It does not ban it in general. It places requirements on high-risk uses, such as transparency and human oversight, and prohibits certain specific practices. Ask a legal advisor about your use.
Related terms
AI conversations, human decisions
Passisto holds structured voice conversations with candidates, employees, clients and students, and turns them into reports your team can act on.