The use of artificial intelligence (AI) has undeniably become pervasive in various sectors. However, when AI is employed to make decisions that profoundly impact people's lives, the demand for transparency becomes paramount.
AI in Welfare Claims
The Guardian highlights in a report that the Department for Work and Pensions (DWP) has increasingly relied on machine-learning algorithms to detect fraud and inaccuracies in universal credit (UC) claims over the last two years.
The use of artificial intelligence (AI) in the welfare system holds the potential for increased efficiency and cost reductions. Nonetheless, it raises serious concerns about fairness, accountability, and transparency.
Secrecy Shrouds the AI System
The government's approach to the AI system has been characterized by secrecy, drawing sharp criticism from transparency advocates.
Big Brother Watch, a transparency campaign group, has described the government's actions as "seriously concerning."
Furthermore, in a prior report, the campaign group pointed out that 540,000 individuals applying for benefits had their fraud risk scores assigned by algorithms in a secretive manner before they could receive support.
At the same time, commercial algorithms process personal data from 1.6 million people residing in social housing to forecast those who may not pay their rent.
The DWP has refused freedom of information requests and blocked MPs' inquiries, citing fears that releasing information could aid fraudsters.
Child poverty campaigners have also voiced their concerns, highlighting the potentially devastating impact on children if welfare benefits are suspended due to AI-driven decisions.
Warning from the Information Commissioner
Information Commissioner John Edwards has warned the DWP sternly, stating that it could face contempt of court unless it changes its approach. Edwards has demanded that the DWP clearly outline the terms under which it could release more information within 35 days.
This development comes after The Guardian's request in July 2022 for information about the AI algorithm's inner workings, the companies involved, and the results of a "fairness analysis" on its impacts.
Shifting Reasons for Secrecy
The DWP's reasons for withholding information have evolved over time. Initially, they argued that releasing data would harm crime prevention efforts and prejudice commercial interests.
Later, in June, they claimed that the cost of gathering information, given the volume of data on the AI system held on government computers, would be prohibitively high.
Amid the secrecy surrounding the AI system, concerns about bias have emerged. The DWP's own trials have revealed evidence of bias in the AI system, raising questions about its fairness and potential discriminatory impacts.
Impact on Vulnerable Populations
Critics argue that while reducing fraud in the benefits system is essential, the use of AI tools has real-life implications for families in poverty who may be caught up in its deployment.
Claire Hall, the head of strategic litigation at Child Poverty Action Group, emphasizes that "The DWP must drop its culture of secrecy and provide meaningful reassurance of the fairness of these tools."
Silkie Carlo, the director of Big Brother Watch, notes that "government uses of AI should trigger much greater public transparency, not less," emphasizing people's right to understand how their information is being used and the rationale behind decisions that affect them.
Stay posted here at Tech Times.
