Artificial intelligence is no longer an experimental add-on in HR: it appears across recruitment and selection, performance evaluation, retention efforts and as support for managerial decisions. The technology can accelerate HR operations and strengthen data-driven decision-making, but it also raises legal, ethical and leadership questions.
This article draws on interviews with Détári István, head of the Edutus University Artificial Intelligence Centre, and Poór József, professor emeritus and head of the Management and HR Research Group at the Hungarian University of Agriculture and Life Sciences.
Why this is a paradigm shift
Poór József views the development as a genuine turning point: earlier major changes in HR were rarely technological in nature, whereas AI can fundamentally alter the role of HR professionals. Détári István also calls it a paradigm shift: although AI research dates back to the 1950s, the current breakthrough is the result of combined factors — increasing computational power, massive datasets, new model architectures and advanced infrastructure.
The shift affects more than HR. Marc Benioff, CEO of Salesforce, has suggested that today’s managerial generation may be the last to lead only human workers; future leaders will manage both people and autonomous AI agents. HR must adapt to that reality.
Where AI can be used responsibly today
Détári notes that AI applications are already widespread in recruitment and selection: screening, questionnaire-based assessments and video interview analysis are common. AI can detect patterns and behavioural cues that are difficult to spot manually. However, capability is not the same as responsible use. The EU AI Act classifies certain AI systems used in employment — for candidate screening or employee-related decisions — as high-risk. The detailed high-risk requirements apply from December 2027, but data protection, human oversight, transparency and documentation are already essential.
Poór offers practical illustrations: a joint Boston Consulting Group and Harvard study that examined nearly 800 consultants found that where tasks matched AI strengths, participants worked faster and produced higher-quality outputs. But there were instances in which AI steered users in the wrong direction and its persuasive phrasing made mistakes less noticeable. Real-world examples also include Amazon’s earlier hiring algorithm, which could disadvantage female applicants due to biased training data, and IBM’s predictive HR analytics that signalled employees at higher risk of leaving — a signal intended to prompt managerial intervention, not automatic action.
Both experts stress that AI can be a powerful decision-support tool, but responsibility for decisions affecting people must remain with humans.
When is an HR-AI system reliable?
Reliability is not merely technical. Traditional enterprise software operates on predefined rules; generative AI produces probabilistic outputs and can give different, sometimes incorrect, answers to the same prompt. That unpredictability has legal, ethical and managerial consequences in corporate environments.
Regulation already limits some categories of application today — for example, workplace emotion recognition is generally restricted.
What skills will future HR leaders need?
Both experts argue that AI literacy will become a baseline competency for HR leaders. That does not mean everyone must become a programmer; rather, HR professionals need an understanding of what tools are for, where they help and where caution is required.
Poór compares AI literacy to learning a foreign language: once you grasp the basics, you better sense the logic of other languages. Détári adds that HR leaders will need more complex knowledge: beyond traditional HR tasks they must act as change managers, understanding how new technologies affect employees, job content, competencies and organisational culture. HR must plan not only how many people an organisation needs but also how work is distributed among people, AI agents and external partners.
Organisational and regulatory challenges
Implementing AI is not just an IT project; it involves business, organisational and HR-process choices, communication and built-in human controls. Poór emphasises that HR has long been a measurable, structured and auditable function, but AI increases the burden of responsibility: organisations must be explicit about which decisions AI supports, what data feeds systems, who supervises them and where human judgment remains decisive.
The macro-level effects are still hard to measure: we cannot yet say precisely how much productivity gain AI will deliver or how deeply it will reshape labour markets. As with earlier technological waves, some jobs will disappear and new ones will appear, while the content of work and required skills will change rapidly.
Education and preparation
The article’s publication was supported by Edutus University, which highlights that navigating AI-driven transformation requires addressing organisational, legal-ethical and leadership issues. The university offers an ‘Data-driven and AI-based HR manager’ programme that integrates data-driven HR, AI decision-support, regulation and change management.
Détári points out that smaller firms often adopt AI tools faster than large corporations, which can be slowed by policies, IT and cybersecurity requirements. The crucial question is how the tools are used: will they become surveillance systems that erode trust or beneficial supports that flag training needs and prompt managerial conversations?
Conclusions
AI promises significant benefits for HR in terms of speed and data-informed decisions, but it also brings new responsibilities and risks. Experts agree that AI should be treated as decision-support: it can inform and accelerate processes, but final responsibility for hiring, retention and dismissal must rest with humans. Organisations and HR leaders should act now — through governance, process design and training — to ensure AI is integrated responsibly and effectively.
