Phai Labs releases whitepaper on building fairer AI for candidate screening
MELBOURNE, Australia, April 6, 2022 /PRNewswire/ -- Phai Labs, the R&D arm of PredictiveHire, has released the whitepaper that informed its FAIRTM Framework, showing how bias in AI can be mitigated with the right steps when using AI for candidate screening.
The FAIRTM Framework (short for Fair AI for Recruitment) received global recognition for its contribution to bias mitigation at the CogX awards.
Phai Labs is releasing the research to further the adoption of their framework and to generate broader interest in the opportunity that AI offers in mitigating bias in hiring.
"Initially my team and I set out to address the fact that badly designed AI systems can amplify bias in hiring, and in doing that, built a framework to monitor and mitigate it," Chief Data Scientist Dr Buddhi Jayatilleke said.
"This research explores the properties that a machine learning system needs to demonstrate to establish a fairer process for job candidates."
Based on the findings, Phai Labs identifies bias, validity, explainability and inclusivity as key measurable properties that AI tools in recruitment need to demonstrate.
These four aspects form the basis of fairness in AI-based recruitment tools, helping developers, employers and candidates ask the right questions when assessing or creating assessment tools.
"This whitepaper consolidates a lot of the current thinking around AI and bias, and we want to encourage as many people as possible to read it because this is such a new area of research," PredictiveHire CEO Barb Hyman said. "Everyone working in this space is on a journey together and it's important we share and learn from each other to build a better future through AI. Our whitepaper shows how mitigating bias with AI is possible, and that is definitely something that more people need to understand. We want more people to know how we can do this so it becomes the standard approach."
Dr Jayatilleke said given this area of research is so new and fast-evolving, it was important his team shared their work broadly so they could get feedback from the community of AI practitioners and researchers. "We are inviting comments from other researchers working in this area. We want to make sure that our framework stands up to robust scientific inquiry."
The whitepaper can be found on ResearchGate:
http://dx.doi.org/10.13140/RG.2.2.17622.73288
About PredictiveHire
PredictiveHire's mission is to help companies unlock and engage talent at scale. Using the world's first Smart Interviewer, powered by the world's largest source of 1st party proprietary text data and advanced Natural Language Processing, we turn simple text conversations into unprecedented talent intelligence enabling organisations to interrupt hiring bias at scale, get to the right talent fast and give every candidate an experience they love.
Share this article