AI in HR put to the test: where it works, where it doesn't
We examine the current role of artificial intelligence (AI) in HR and conclude that its potential remains untapped in many companies – despite growing openness. Successful practical examples such as the SAP chatbot “Joule” and AI-supported skill management show how AI can make HR processes more efficient and objective. At the same time, data protection concerns, a lack of knowledge, unclear costs, and poor data quality are slowing down widespread adoption. To close the gap between potential and reality, we recommend a strategic, step-by-step approach with proven applications, a good database, and clear communication.


AI – A Status Report
AI is gaining momentum
AI is no longer in its infancy – and has the potential to change HR forever. However, this reality has not yet reached the market, as the Digital HR 2024 study shows. According to the study, most companies are completely unaware of the possibilities offered by AI. Only a few are testing individual AI functionalities, and virtually no companies are using AI across the board. At the same time, two out of three respondents believe that now is the right time to push ahead with AI in HR. However, very few are taking action. Where does this gap between expectation and implementation come from?
AI in HR: Hype does not equal benefit
The latest Gartner Hype Cycle for HR Technology sheds light on the situation. It shows how individual AI applications have different degrees of maturity and are therefore perceived and evaluated differently. For example, companies have high expectations of AI-supported skill management – an approach that uses AI to identify and analyze employees' skills in order to derive tailored recommendations for further training. At the same time, it is not yet entirely clear what opportunities this opens up for skill management as a whole.
The situation is different when it comes to the use of generative AI, for example in assistance tools. These are already widely used and make recruiting and talent acquisition more efficient and easier, as the following application examples show.
Jan Bitterli, Junior SAP Consultant
AI applications make HR processes more efficient and simpler. The best examples are applications in recruiting and talent acquisition.
Joule: the multi-talented recruiting tool
Perhaps the AI application par excellence at the moment is Joule, the chatbot in SAP SuccessFactors. It has recently become available in Microsoft Teams. This makes the recruiting process even faster and easier, as HR managers and executives can now work in MS Teams and don't have to switch to SAP SuccessFactors. Joule not only reminds managers about job interviews, but also creates the appropriate interview questions as needed.
The two applications “Payroll Statement” and “Position Creation” demonstrate that Joule can do even more. For example, employees can view their pay slips with just a few clicks – at any time and for any period. This saves them time and reduces the workload for the HR team. The same applies to position creation. What used to require tedious coordination loops between HR and managers can now be done in seconds. This speeds up the recruiting process immensely and keeps it lean.
Skill Management: A prime example
AI is just as helpful in talent and skill matching as it is in the applications mentioned above. The key feature is that AI checks the extent to which the skills listed by applicants match the requirements of the job profile. The system displays the candidates with the best matches, but also shows which skills they still lack. This significantly speeds up the selection process and, in the best case, makes it more objective by compensating for and eliminating distortions caused by subjective perceptions.
What is slowing down AI
As clear as the advantages of AI are, so too are the concerns of companies regarding its widespread use in HR. We also notice these reservations in our discussions with customers. They include:
- Reservations due to data protection. Since HR processes personal data, data protection applies. To what extent is this guaranteed when using AI? What are the consequences of violations? These are questions that need to be answered, especially given the different approaches and regulations in Europe and the US.
- Reservations due to a lack of transparency. The frequent absence of clear control mechanisms in AI applications creates mistrust. This is also because the algorithms behind AI remain hidden. In addition, many people fear for their jobs as a result of the use of AI.
- Reservations due to a lack of skills. Not knowing also breeds mistrust. Surveys and studies show that AI is unfamiliar to the majority of the workforce. As a result, they do not know how to deal with it.
- Reservations due to insufficient data quality. The benefits of AI stand and fall with the quality of the underlying data. The better the data, the greater the advantages. This means that investing in AI applications only pays off if the data quality is right and distortions can be ruled out (see Skill Management and Recruiting).
- Reservations due to unclear costs. Many companies want to invest in AI applications. However, the costs are not tangible for many. A plausible cost-benefit analysis is difficult because the actual benefits only become apparent later.
Sebastian Krebs, Managing Consultant
The hype surrounding AI has given way to a development that will also bring about lasting change in HR. Companies should act now.
Develop a strategy
The order of the day is to act, set clear goals, and create a roadmap. But which measures should be prioritized? And what are the long-term goals? We recommend that companies start with tried-and-tested AI applications. These include, for example, AI-supported job advertisement creation and automated resume analysis, known as CV screening.
Such applications deliver rapid added value, which in turn helps to build acceptance and trust. This effect can be further enhanced by communicating the introduction of the software modules honestly and transparently and involving key stakeholders in the process.
As a basis for further AI expansion, data quality must be right and clear AI governance must be established. A central component of this is ethical guidelines that answer questions such as: Which decisions can be made by AI in the company? How is the necessary objectivity ensured? Etc. At the same time, it is important to establish training programs to make HR fit for AI. This will open up completely new perspectives for HR, resulting in better HR processes and associated new tasks, roles, and structures. Now is the time to get started.