A Judicial Wake-Up Call on Government by AI
A UK judge blocked the Home Office's use of an AI lie-detection system at borders, ruling its video analysis was not objective evidence. The decision signals growing judicial skepticism toward automated government decisions lacking transparency and legal authority.
Background
- A federal judge in the US ruled that a government algorithm used to determine disability benefits (for Social Security Disability Insurance, SSDI) likely violates the Administrative Procedure Act (APA), which governs how federal agencies make rules and decisions.
- The case, *Gillespie v. SSA*, challenges the Social Security Administration's use of an AI system to assess claimants' "residual functional capacity" — essentially, what work they can still do despite disabilities.
- The ruling is significant because it suggests that when agencies use AI to make individual determinations, they must follow the same legal standards as human decision-makers — including providing meaningful reasoning and allowing claimants to challenge the evidence.
- This is part of a broader legal and policy debate: as governments increasingly automate welfare, benefits, and even criminal justice decisions, courts are starting to ask whether existing law permits "government by algorithm" without proper transparency, oversight, or due process.