One Critical Limitation of AI in HR

Today more than ever, people want to do meaningful work, and organizations want to maintain their competitive advantage by attracting and retaining the best talent. AI is seen by many as helping them with this vital task.

AI Technology in HR

AI has shown tremendous potential for solving HR problems, including; taking care of service tickets, producing reports, and ensuring data is stored and disseminated accurately without needing human intervention. According to PWC research, 40 percent of the HR-functions in international companies are currently using AI-applications in some form. Most of them are still US-based, with European and Asian organizations trailing behind.

However, there are limitations. It is widely agreed that there are some things AI can’t replace, such as human creativity, emotional intelligence, and empathy, to name a few.

It’s interesting to note that the spectrum of enthusiasm for new HR technologies closely follows the technology adoption lifecycle; early adopters fully embrace AI and all that it has to offer, and then there are skeptics who are more cautious about it.

Limitations of AI in HR

In the HR spectrum, there is one thing that rarely gets the attention and consideration that other HR functions do: the ability to, and the importance of, accurately paying someone. Understanding this process is crucial given the changes regularly happening from a compliance and regulation standpoint, but also because of its morale and employee satisfaction implications; no one wants to be paid inaccurately, or not paid at all.

Discover the four key principles of managing payroll using your HRMS with our free guide

What happens to employee morale when things go wrong?

In a recent example of AI gone wrong, an employee was fired by an AI driven process and it took weeks to resolve. In the meantime, the employee went unpaid and eventually left the company due to the feelings created during this process. There needs to be quick correction and communication with employees when the process fails to deliver the intended result to avoid negative outcomes.

Portions of accurate pay can and should be automated, but the process still requires human verification and approval. Accurate pay and employee data also become the basis of the rules for AI driven business process automation.

HR professionals have plenty of new tools and technology at their fingertips. The goal of any of these technologies is to free up HR and Payroll professionals from redundant and repetitive tasks so they can focus on more strategic ones. There is more work for HR and Payroll than ever before: compliance and regulation laws are constantly evolving, and increases in contingent and contract workers mean time tracking and payroll needs are more complicated too, possibly requiring cross border payments depending on where contractors reside. The tools do help, but detailed, accurate data is required to realize their full benefit, including generating reports, and analytics.

 It is true that AI is extremely good for repetitive tasks. AI driven processes can replace workers doing routine, methodical tasks, which helps to better utilize a human’s ability for problem solving, leadership, emotional intelligence, empathy, and creativity.

In many organizations, there is still improvement to be made in accurately paying and effectively managing employee records as well as the overall employment experience. AI is in its early stages in organizations and still requires careful review to make sure that employees still feel connected to the organization and their needs are met when processes are automated. None of us want our very valuable (and difficult to recruit) workforce to leave the organization because of bad feelings or mistakes made by AI.