Getting started with enterprise AI governance

In some business settings such as finance and healthcare, the machine learning algorithms that underpin AI are already regulated. For example, in the UK, the Financial Conduct Authority (FCA) is responsible for ensuring that credit agencies who use risk modeling (typically logistic regression) to estimate the likelihood of a customer defaulting on a loan do so in ways that are consistently fair to the customer. The credit agency must assure the FCA that the machine learning models they use for this are not biased for certain customer groups, are robust estimators of default risk, and similar assurances.

So, we can say that in some instances, AI Assurance Frameworks already exist and are used regularly in business today. However, as the use of AI grows exponentially across industry and the machine learning models used to produce this AI increase in their power and complexity, there is arguably a growing requirement for AI Assurance in...