Does the EU AI Act allow talent acquisition technology to balance algorithmic efficiency with legal compliance?

We recently published an in-depth article about whether laws such as the European Union Artificial Intelligence Act make the scoring, ranking, or matching of candidates against job openings illegal.

Today, I want to bring to you the perspective from a lawyer, but not one whose practice is focused on the EU, AI, job boards, recruitment, or any of the other usual suspects. Matthew Pfau, is an attorney and the founder of the Paralegal Institute. As a law firm owner who utilizes AI-driven research tools and objective “Paralegal Scorecards” for hiring, he has first-hand experience balancing algorithmic efficiency with legal compliance.

According to Pfau, “The EU AI Act does not ban candidate ranking but categorizes recruitment AI as ‘high-risk,’ meaning these systems are illegal only if they lack human oversight or fail to meet strict data quality standards. If a recruiter uses a tool like LexisNexis for skills testing or an ATS for scoring, they must provide a clear mechanism for human intervention to remain compliant.”

Pfau is no luddite. He states that, in his practice, he has “In my practice, I’ve “found that while technology like automated checklists can significantly improve workflow, the law requires that these tools do not replace human judgment in critical career decisions. These systems face illegality when they function without the rigorous risk-management frameworks required to protect an individual’s right to a fair evaluation.”