Artificial Intelligence in Human Resources: Essential Guidelines for Employers and HR Professionals

The use of artificial intelligence systems in recruitment and human resources management is no longer unregulated. Regulation (EU) 2024/1689 on artificial intelligence (“AI Act”) establishes clear rules in this regard, and compliance is essential for employers and HR professionals.

The regulation entered into force on August 1, 2024, and will be implemented in phases.

• As of February 2, 2025: Prohibitions on AI practices deemed unacceptable are already in effect.
• As of August 2, 2026: The strict set of rules regarding high-risk AI systems, along with the applicable penalties for companies, becomes fully enforceable.

High-Risk AI Systems

According to Article 6 and Annex III of the regulation, software applications used for the following purposes are considered “high-risk” systems:

Automated CV screening;
Candidate assessment and testing;
Promotion decisions;
Monitoring workplace performance and behavior;
Any decisions that directly affect employment relationships.
These systems can directly affect the fundamental rights, access to employment, and career advancement of employees or candidates.

Strictly Prohibited Practices

Article 5 of the Regulation categorically prohibits certain practices that violate human dignity:

Manipulation of individuals’ behavior;
Exploitation of vulnerabilities (age, disabilities, socioeconomic status);
Social scoring.
Excessive or inappropriate use of behavioral or emotion-recognition analyses during interviews creates significant legal risks of discrimination.

Key Obligations for Employers

If you use high-risk AI systems, the AI Act requires you to ensure:

Full transparency: Job applicants and employees must be aware that they are interacting with an AI system.
Human oversight: No major final decision may be left solely to the algorithm.
Anti-bias testing: Periodic verification of systems to prevent and reduce discrimination.
High-quality data: Use of accurate, relevant, and representative datasets.
AI literacy (Article 4): Companies are required to train and develop the skills of staff who operate these AI systems.
Applicable legislation in Romania

The AI Act is not a standalone piece of legislation but rather complements an existing national and European legal framework. The use of AI in HR must be aligned with:

GDPR (Regulation (EU) 2016/679) – In particular, Article 22, which prohibits decisions made solely through automated processing if they produce significant legal effects on the individual;
Labor Code – Law No. 53 of January 24, 2003;
Law No. 190/2018 (on measures for the implementation of the GDPR).

Risks and Penalties

Failure to comply with the AI Act may result in fines of up to 35 million euros or 7% of annual global revenue.

To mitigate legal and reputational risks, companies must:

1. Implement clear internal policies governing the use of AI.
2. Conduct rigorous due diligence processes for all AI solution providers.

Conclusion

The AI Act profoundly changes how companies can use artificial intelligence in human resources. The technology must remain a support tool, not an absolute decision-maker, and employers and HR professionals must give absolute priority to transparency, human oversight, and the protection of personal data.

In this new legislative landscape, implementing clear internal policies and carefully vetting the AI solutions used are essential steps for avoiding major legal, financial, and reputational risks.