A new survey conducted by ResumeBuilder.com shows what could be a significant trend toward the use of artificial intelligence in human resources. Artificial intelligence (AI) increasingly plays a central role in informing those decisions. The survey interviewed 1,342 HR managers. It sheds light on their increasing use of AI tools to determine or inform major decisions about personnel—promotions, raises, layoffs.
Those findings indicate that a staggering 78% of managers relied on AI to determine whether or not employees deserved a raise. Moreover, 77% used AI to help them decide whether to promote employees. This trend further highlights an inflection point towards data-driven decision-making in HR departments. This dependence on technology raises profound ethical issues. Almost half (49 percent) of these surveyed managers have been prompted by leadership to think about whether or not AI could replace the jobs of their direct reports.
Surprisingly, one HR manager in five openly acknowledges that he allows AI to decide who gets laid off. This deeply disturbing trend all happens without any human intervention. An increasing share of HR managers are adopting this practice. In fact, it’s 60% that are using AI tools just to help them get through the layoff process. Additionally, 66% of these professionals admitted that they ditched human oversight entirely during the layoff process by using AI to do so. Within this niche, 20% said that they relied on AI to actually decide which employees to lay off.
Although these statistics are a positive sign about the increasing adoption of AI in HR processes, experts caution against some of the risks involved. Soribel Feliz currently directs AI governance and regulatory compliance at the U.S. Department of Homeland Security. She emphasized an important area of need for training HR professionals on the ethical use of AI.
“Most managers lack formal training on ethical AI use in HR.” – Soribel Feliz
This phenomenon, in AI discourse often referred to as the “LLM sycophancy problem,” casts doubt on the robustness of AI in these scenarios. This phenomenon, known as confirmation bias, is when generative AI reflects and amplifies a user’s pre-established viewpoints, often prejudicing all decisions made with the tool thereafter.
Despite these challenges, AI tools such as ChatGPT have become fixtures in people management, streamlining processes and providing insights that were previously unattainable. The survey findings reveal an overwhelming demand for improved AI governance in HR. Few if any have developed or deployed extensive ethical frameworks for practices.
