AI hiring inspected
Artificial intelligence could help reduce bias in the hiring of workers.
A new artificial intelligence-driven platform has been assessed against key ethical hiring criteria of transparency, privacy, bias and accountability.
Experts behind the project sought to create a ‘non-biased talent shortlisting algorithm validation’ scheme - a practical exploration of ethical AI.
“You are dealing with sensitive information about real people, so building trust into that process is critical,” says project leader Professor Fang Chen, Executive Director Data Science at the University of Technology Sydney.
“AI for good needs to be the standard. But there has been no way to properly assess that until this project.”
Over the last two years, the research team lead by Professor Chen has developed, tested and iterated an assessment process before its use by industry partners.
She says this process is needed to confirm that the AI outputs are fit for purpose and deliver actionable results.
The benefits that data and AI are bringing to the professional workforce are phenomenal, but AI is not immune to bias in the data or in the algorithms.
Previously, the decision making has been hidden, meaning until now, there has been no clear, defensible, independent, and objective validation demonstrating ethical AI.
Mark Caine, Artificial Intelligence and Machine Learning Lead, World Economic Forum said it is imperative to minimise the risk of AI to humanity, otherwise the public will lose trust in AI and its capability to do good.
While there are over 200 AI ethics frameworks and guidelines globally, few have been operationalised.
The experts say the new UTS project is a milestone in bringing independently audited certification to an AI product.
They predict AI will offer automated matching of potential candidates or employees to opportunities in a way that removes negative unconscious bias from the process. It should also be able to help HR users to explain why talent has been recommended to ensure it complies with Equal Opportunities and employment law.