Artificial intelligence is increasingly finding its way into human resources departments. Decisions about promotions or terminations are no longer made only by people. What initially seems efficient raises profound ethical questions.
Where AI already judges people today
The integration of AI into corporate personnel decisions is no longer a vision of the future. Some companies around the world, including Europe, are using artificial intelligence to evaluate employee performance, recommend promotions and even terminate employees. In a study by the US platform Resume Builder, 1,342 managers were surveyed about their use of artificial intelligence in their everyday work. The results are clear: the majority said they use Large Language Models (LLM) like ChatGPT or Gemini to answer big HR questions. While 78 percent use language models to decide salary increases, 77 percent use LLMs to make promotion decisions.
But AI tools are not only used for positive decisions. When it comes to terminations, just over 60 percent said they use AI. But that’s not all: More than one in five respondents said they let an AI make the final decision without human reassurance – a process that has profound ethical and legal implications.
Between efficiency and dehumanization
The arguments for using AI in human resources management are understandable at first glance, because data-based decisions are intended to promote fairness and efficiency. But the reality is often different. In practice, many employees report alienation in the work environment. Termination “by an AI” is often perceived as inhumane. Emotional aspects, individual life situations or internal team dynamics are left behind in algorithmic evaluations. Although AI can process large amounts of data, it cannot adequately evaluate context or human nuances. Incorrect or biased data sets quickly lead to discriminatory results and usually without transparent correction options.
The responsibility lies with people
The increasing automation of personnel decisions through AI requires a social discussion. Clear rules and guidelines seem necessary. Artificial intelligence can support companies, but should never be the sole judge of people’s fates. Decisions about dismissals must always be comprehensible, individually justifiable and humanly responsible. There is a need for responsible interaction between humans and machines, with transparency about the use of AI as well as legally binding limits on their decision-making powers and clear traceability of processes.
Editorial team finanzen.net