AI Risks: A Fluctuated Bidding for Online Jobs’ Ads to Mitigate Gender-Based Discrimination Risk

In the first stage of the AI-based recruiting, AI is used to predict then deliver jobs’ ads to the most relevant audience on online platforms. Empirical studies have proved that AI has discriminated based on gender during the jobs’ ads delivery process and the literature discussed multiple reasons behind it. This work focuses on one important reason which is the mere fact that females are more expensive demographic. As the recruiters’ advertisers are in a competition with other commercial advertisers who bid high to win online ads’ placements for females, this will give little chance for jobs’ ads to be delivered to both genders in an equal distributed manner. This paper proposes a framework that especially designed to conduct an AI-based controlled and supervised bidding process to deliver jobs’ ads in a near equal distribution between males and females separately from the conventional greedy logic behind delivering commercial ads where high bids is used to achieve more revenues with a skewed delivery.