UK warned over lack of transparency on use of AI to vet welfare claims
The UK Department for Work and Pensions faces severe legal and ethical scrutiny for its opaque use of artificial intelligence in welfare fraud detection. The Information Commissioner has warned that the department risks contempt of court by refusing to disclose essential details about its machine-learning algorithms. This lack of transparency prevents public accountability and raises serious concerns about the fairness and accuracy of automated decisions that determine individuals' financial survival. This issue is critically relevant to open data because it highlights the urgent need for algorithmic transparency in public sector decision-making. Without open access to the data, logic, and performance metrics of these AI systems, citizens cannot understand why benefits are suspended or challenge discriminatory outcomes. The department’s refusal to share information undermines the fundamental principle that government data should be accessible to ensure trust and prevent the marginalization of vulnerable populations. Ultimately, the situation underscores that technological efficiency cannot outweigh the right to know how automated systems impact society. Civil liberties groups and child poverty advocates emphasize that opaque AI poses a high risk of unfair and discriminatory results against ethnic minorities, the elderly, and disabled people. True open data requires more than just releasing statistics; it demands clear, accessible explanations of how algorithms work to protect human rights and ensure justice in the welfare system.
Source: theguardian.comPublished on 2023-09-04