Member

The second important point is that, independent of the measures we are introducing in the Bill, the DWP already has meaningful human involvement in how it tackles fraud and error more widely. The DWP currently has one fraud and error machine-learning model in full deployment on universal credit advances. Our use of machine learning in fraud and error incorporates human intervention to consider the case, make further inquiries if necessary and then make a decision based on all the evidence. There is no automated decision-making when investigating fraud and error, and our use of advanced analytics does not replace human judgment. To provide further assurances to Parliament and the public about our use of AI and our processes, the DWP intends to publish fairness analysis assessments that will set out the rationale for why we judge our models to be reasonable and proportionate.