The EU AI Act isn’t just coming for your TA tech algorithm, but also your training datasets

We published an article a few weeks ago about the impact of the European Union Artificial Intelligence Act on job boards, applicant tracking systems, and other talent acquisition technology that ranks, scores, and matches candidates against job openings. Today, we’re going to shift away from discussing the algorithms and toward the data used by these systems.

We reached out to Hansjan Kamerling, who is a product designer and the founder of Adaptify AI. He’s built search experiences for 33 marketplaces used by millions and managed custom-trained AI models for 40,000 users. His current platform, Adaptify AI, automates high-stakes SEO workflows, which requires him to bridge the gap between automated ranking algorithms and strict regulatory standards.

According to Hansjan, “These systems become illegal if they violate ‘data governance’ mandates, which require training datasets to be representative and free of biased patterns. Unvetted data can mirror historical prejudices; under the new Act, any scoring tool using such data to rank candidates is strictly non-compliant.”