The increasing reliance on artificial intelligence and digital technologies in the UK government's social security system has led to the systematic elimination of people with disabilities, low-income communities and other marginalized groups, warning Amnesty International on Thursday in a new report. The report, entitled “Too many technology and insufficient empathy,” sheds light on the digitalization of the work and pension sector (DWP) of major welfare schemes such as Universal Credit (UC) and Individual Independent Payments (PIP).
According to the Amnesty International research findings, the deployment and frequent revisions of AI-based systems create highly inaccessible environments for those who need welfare the most. Many applicants lack digital literacy, internet access, or compatible devices, are trapped in bureaucratic scope and become even more impoverished by delayed or denial of benefits. Phone alternatives often suffer from long wait times and limited assistance.
The organization called on the UK government to conduct an independent and impartial review of DWP's digital systems, to eliminate technologies that violate human rights obligations, and to ensure that AI systems are not made mandatory. The organization highlighted the urgent need for AI laws centered around transparency, equity and human surveillance.
The study was based on Amnesty's May 2025 report, “Social Anxiety: The Catastrophic Human Rights Impact of the UK's Social Security System Failure,” and called for a complete overhaul of the UK's benefits system. The two reports collectively demonstrate that, based on data from 782 interviews with claimants and welfare advisors, digital systems exacerbate existing inequalities rather than mitigating them.
Critics say the government's enthusiasm for cutting costs through automation overshadowed its obligation to protect dignity and rights. Kiel Prime Minister Starmer plans to accelerate AI across the public sector by making Britain an “AI superpower.” “We've seen the best of our efforts to help you understand how we can help you,” said Bishop Imogen Richmond, an international researcher on technology, economic, social and cultural rights.
DWP's mission to reduce “costs” is an unhealthy obsession and an overreliance on flawed technologies. But urgent questions remain. Is technology deployment really cost-effective, or is it just cutting corners at the expense of vulnerable people? …DWP experiments using Tech Systems have reduced people to data points. They oversimplify people's complex reality by creating narrow, opaque processes that undermine people's needs, especially when they can't get support from the human caseworkers they need.
The UK government has previously touted digital transformation as a way to improve efficiency and accessibility. However, critics argue that these efforts failed to explain digital poverty and living experiences. In a recent UK government report on early adoption of AI in education, the agency admitted that “the biggest risk is nothing,” but that same logic appears to ignore the risk of doing wrong things without accountability.
Other international watchdogs have also flagged concerns. In a 2022 report to the Human Rights Council, the United Nations High Commissioner for Human Rights warned about the digitalisation of welfare systems, saying that despite its potential benefits, these systems risk eliminating those most needy and underscore the need for more robust protection in algorithmic welfare programs. Similarly, in Denmark, AI-driven social support systems have been found to disproportionately punish marginalized groups and fuel mass monitoring.
Rights advocates emphasize that technological advancements should not be at the expense of human dignity, equity and access. Without meaningful reforms, these advocates suggest that the current system could continue to deepen digital division and undermine the rights of the very people designed to support it.
