Today we continue our conversation around risks of bias in AI. We look at how architects can address the risk of AI models adopting bias by regularly maintaining and auditing AI models for any sort of fault or weakness in their decision making. We talk about the importance of regularly updating AI models and how it can address any sort of bias inherent in the model as well as help improve its effectiveness in its design.

AI is like any other tool we use, it needs to be maintained regularly for it to be effective. Additionally, like any other tool we use, there has to be a human element in any sort of decision making process, especially with important decisions that can have a direct impact on people's lives, like medical diagnoses and loan applications.