Brian McElligott

However, almost the same amount (64 per cent) said that they are using AI in their HR practices to some degree.

Catherine O’Flynn, employment partner at Mason Hayes & Curran, said: “The disconnect between AI use and governance could expose companies to significant legal and reputational risks. With HR AI systems now classified as ‘high-risk’, aligning practices with regulatory requirements is crucial. Companies that proactively address these challenges will be better positioned to leverage AI’s benefits while minimising legal risks.”

The survey identified data protection as the biggest challenge in complying with the EU AI Act in HR (43 per cent), followed by regulatory clarity (29 per cent) and AI bias (28 per cent). These concerns align with the Act’s stringent requirements for high-risk AI systems, which include obligations for data governance, transparency, and human oversight.

Brian McElligott, partner and head of AI at Mason Hayes & Curran, said: “Data protection remains a central issue, tied closely to existing GDPR obligations. The AI Act introduces new layers of complexity for high-risk AI systems used in HR. Businesses need to prioritise risk assessments, create detailed AI governance policies, and ensure their workforce is adequately trained on AI’s regulatory framework.”

Recruitment emerged as the area where AI could deliver the greatest benefits (41 per cent), followed by work allocation (32 per cent) and performance management (27 per cent). However, opinions on operational efficiencies gained from AI in HR were split, with 50 per cent of users reporting benefits and 50 per cent seeing no improvements.

Ms O’Flynn added: “It’s not unexpected to see mixed results at this stage of AI implementation. Realising the full potential of AI in HR requires not only careful planning but also robust data management and ongoing monitoring. Businesses must invest in these areas, particularly given the new compliance obligations imposed by the AI Act.”