PayPal 'Fairness-Aware' Model Patent Highlights Bias Issues in Financial AI
PayPal is planning on patenting a system to adjust training data sets for AI. They have created a sampling scoring method to create more fair datasets that they will train models on, and they hope that their method can be extended to other aspects of the training process.
“Most people take the approach of, ‘we can’t do ethics if we don’t have the AI in the first place,’ so they very much concentrate on AI,” said Green. “But sometimes (ethics) comes as an afterthought, which is always a risky approach to it.”
Green also commented on the trajectory of ethics in AI: “It’s good to see companies filing these sorts of patents that show they’re doing the work. It sends an important signal that they find it to be worthwhile work.”
Brian Green, director, technology ethics, quoted by The Daily Upside.