FairByDesign Reporting
automated decision-making
All reportingInvestigationsMercury BriefField papersGovernance resourcesNews
-

Is “Human Oversight” a Useful Fiction?
Your name goes on the approval, and if the AI was wrong, the blame may land on you. Two strong arguments say human oversight is a fiction: one says scrap…
-

1.2 Seconds per Claim: Where Should AI Need a Real Human Yes?
A doctor’s name on a claim denial looks like human review. ProPublica reported Cigna doctors averaged 1.2 seconds per claim; Cigna disputes the account. Here is what makes a human…
-

The Bank Chatbot Sounds Sure. Here’s What It Can’t Promise You
The chat window in your banking app is fast, friendly and confident. It is not where your rights live, and a chat message may not even count as notice of…
-

Episode 1: Is your AI really just a suggestion?
A human holds the final button, so the AI is “just a suggestion”? In the first episode of the Operational AI Readiness show, Viktoria Bakos uses a real child welfare…
-

Human Presence Is Not Human Authority
A human somewhere in the workflow does not mean a human is in control. What the law actually asks of “human oversight,” what the research says about whether people can…
-

The Human Witness
The final part of The Evidence Chain. Robodebt treated a calculation as proof of debt; old payslips settled the truth, in both directions. As machines mediate more of what we…
-

The Correlated Decision Problem
Part 6 of The Evidence Chain. He applied at 12:55 a.m. and was rejected by 1:50. When many employers and lenders read you through the same model, a hundred rejections…
-
The AI Enforcer Already Open for Business Is Called the GDPR
While much of the AI Act’s use-based high-risk regime is set to move to December 2027, and product-integrated high-risk systems to August 2028, data-protection authorities have kept fining AI-enabled biometrics…