Recommendation 22

As part of our previous examination of the Department’s 2019–20 Accounts, we recommended that the Department should monitor and report any discrimination or bias caused by using artificial intelligence and machine learning on different claimant groups.41 In its response to our report, the Department agreed with our recommendation and told us that it had a Data Science Ethics Framework for machine learning that ensures it considers bias and discrimination in the design of predictive models. It also told us that it was working with legal experts to ensure that the ethical position of its products have been properly considered ahead of any wider automation, and committed to providing an update on how it is using data to tackle loss as part of its Annual Report and Accounts.42 The Department told us that wider use of data analytics methods such as machine learning will be a key part of its future efforts to prevent fraud from entering the benefit system. During 2021–22 the Department trialled a machine learning model that uses information from past fraud cases to build a prediction of which claims are likely to be fraudulent in the future.43