Could artificial intelligence (AI) remove bias from the hiring process? A team of researchers at Case Western Reserve University wants to find out.

With a one-year, $250,000 grant from Walmart, the team aims to address critical challenges in the current environment of so-called learning and employment records (LERs) in human resources.

LERs are essentially digital resumes with verified records of people’s skills, educational experiences, and employment histories.

“The current learning and employment records systems are overly complex and disjointed, making it difficult for people to clearly and credibly demonstrate all of their skills to employers,” said Youngjin Yoo, one of the study’s principal investigators, associate dean of research and the Elizabeth M. and William C. Treuhaft Chair in Entrepreneurship at CWRU’s Weatherhead School of Management.

Yoo found that the current AI-powered hiring process often suffers from bias due to a lack of valid and reliable data that indicates people’s true skills and competencies, leading to poor hiring outcomes and skills gaps.

Researchers believe that optimized AI could provide a solution by improving the hiring process and more accurately matching skills with job requirements.

But integrating AI into the LER ecosystem also brings challenges – such as privacy, fairness, transparency and auditability. Centralized LER systems often raise concerns about privacy and security, and individuals typically do not have control over their personal data, says Erman Ayday, an associate professor of computer science at the Case School of Engineering who is leading the research with Yoo.

To solve these problems, the team proposes developing a decentralized, privacy-preserving data architecture that leverages AI, Ayday said.

“Walmart is excited to support this groundbreaking research,” said Sean Murphy, Director of Retail Opportunity at Walmart. “We believe that improving the hiring process with AI can lead to better outcomes for both employers and job seekers and create a more equitable labor market. People often think that AI is a threat to their jobs. The project will demonstrate that AI can actually help people find better jobs.”

“This collaboration is a significant step toward revolutionizing the way we manage and verify employment records,” said Morgan Frank, assistant professor in the Department of Computer Science and Networked Systems at the University of Pittsburgh, who participated in the research along with the Digital Credential Consortium (DCC) at the Massachusetts Institute of Technology.

“By using privacy-preserving AI to enrich the data contained in digital credentials, we can improve their usefulness for the labor market,” added DCC Director Kerri Lemoie.

The research will focus on several key topics:

• Designing a decentralized data system that gives people control over their personal data while providing them with the ability to securely share their learning and work records.
• Using AI to infer skills and abilities from things like certificates and job descriptions.
• Creating privacy-friendly methods to use AI to match skills to jobs without compromising data privacy.

“The goal is to create a fairer, more efficient and more privacy-preserving infrastructure for LERs,” Yoo said. “This will empower individuals to showcase their skills, enable employers to make more informed hiring decisions, and foster a more dynamic education and employment ecosystem.”

For further information please contact Colin McEwen at [email protected].