Artificial intelligence (AI) and machine learning (ML) are now embedded in our daily lives, whether it’s in the selection of films recommended by your preferred streaming service or a list of suggested featured products to buy on your online shopping carts. These features use algorithms to predict what you are likely to consume based on past behavioral and consumer data.
The use and expansion of AI capabilities have been growing exponentially, with investment in the emerging technology alone increasing 15-fold in five years. This trend suggests that organizations have high expectations for emerging technologies to help maximize output, profitability, labor productivity, and engagement. However, this rapid growth trajectory may also undermine the unintended risks involved in using them.
While the use cases for AI and ML can be very compelling and the risks relatively low when it comes to predicting consumer behavior for goods and services, the stakes become much higher when both are used to drive workforce decisions in the firm.
This is the conundrum that Human Resources organizations are confronted with in this day and age. On one hand, there is a consensus among HR practitioners that emerging technologies such as AI and ML have the potential to help increase productivity, reduce human bias, and improve accuracy in workforce predictions. On the other hand, the latest research suggests that there is also a growing concern about the ethical, legal, and lack of transparency behind AI and ML tools.
The question then becomes – What can CHROs start doing to address these concerns and leverage these emerging technologies effectively? Let’s take a look:
- Increase transparency – Candidates and workers tend to become more skeptical and worried about HR-related decisions driven by AI and ML outcomes when they do not understand the what, who, how, and why. Therefore, it becomes critical for employers to be ready to explain the use cases, algorithms, and decision-making processes. Additionally, adhering to compliance with the different data protection regulations is paramount. Most laws require employers to request candidates and workers to opt-in to data collection, be able to appeal decisions resulted from algorithms, and also maintain confidentiality.
- Review data sets for bias and expand data sources – One of the biggest risks when it comes to AI and ML is limited and potentially biased data sets. Therefore, it’s really important early on that CHROs take a deeper look into their employee data and evaluate areas such as performance, potential, promotions—and conduct a very thorough adverse impact analysis to look for a potential indication of bias. Another risk is the limited availability of data. AI and ML are most effective when used in very large data sets, so exploring the use of other data sources, including from third parties that can expand the use of those algorithms and ensure it’s more accurate.
- Take a multidisciplinary approach – Many organizations are now focusing on hiring data scientists that can build algorithms and mine for data. The complexities of using AI and machine learning in human resources require that CHROs assemble their team of individuals with multidisciplinary skill sets to effectively assess and apply these emerging technologies in organizations. A good way to tackle this challenge is by forming an internal AI council composed of legal and ethics professionals, data scientists, HR practitioners, and workers to review and assess any potential adverse impact on use cases for the aforementioned emerging technologies.
Technology alone cannot address the shortcomings of AI and ML applications in HR. As Ben Shneiderman, a scientist and researcher at the University of Maryland suggests, there is a need for a human-centered AI approach which calls for a more balance between human control and computer automation. In other words, technology should be used to make the workplace more human, not less.
Enjoyed reading this article? Subscribe to Oracle’s modern HR blog.