A Scalable Framework for the Verification of AI Ethics with Multiple Diversity Elements in AI-Informed Talent Decision Making

Project Member(s): Chu, W., Chen, F., Berry, A., Zhou, J.

Funding or Partner Organisation: Reejig Pty Ltd
Reejig Pty Ltd

Start year: 2022

Summary: This project is an extension to the previous REEJIG project in the assurance of AI ethics by adding additional diversity elements in the ethical AI verification framework. In different countries such as US, EU, and Australia, state and national laws have different provisions on equal employment opportunity and anti-discrimination in the workplace. A workplace with diversity and free from discrimination brings various benefits to employers including productivity improvement and work efficiency increasement. Diversity in the workplace means having employees from a wide range of backgrounds. However, unconscious bias because of different factors such as socio-economic factors in resumes and work history, even at the recruitment level, can hamper an organisation's attempt at achieving diversity. With the increasing use of Artificial Intelligence (AI) in talent decision making, the verification of bias in AI solutions is becoming indispensable to allow users to gain insights of biases in AI and make fair decision-making in the talent decision making process. Besides the gender balance issues that most of the current research including the previous REEJIG project focuses on, other diversity issues such as cultural background, ethnicity, age, mobility or impairment, location, and socio-economic status are also major concerns in the workspace diversity. However, little work has been done to verify the unconscious bias regarding these aspects in AI-informed talent decision making. Since various aspects of concerns need to be considered, it is a challenge to consider these ethical concerns at the same time for talent decision making. Furthermore, various AI algorithms have been used in talent decision making and it is valuable to compare which algorithms give better performance in bias comprehensively.

FOR Codes: Artificial intelligence, Social ethics

Acknowledgement of Country

UTS acknowledges the Gadigal people of the Eora Nation, the Boorooberongal people of the Dharug Nation, the Bidiagal people and the Gamaygal people, upon whose ancestral lands our university stands. We would also like to pay respect to the Elders both past and present, acknowledging them as the traditional custodians of knowledge for these lands.