In January 2026, a newspaper investigation asked Google a simple question, "What is the normal range for liver blood tests"? The AI summary at the top of the results answered with a set of numbers. Those numbers did not account for age, sex, ethnicity or nationality, the very factors that change what 'Normal' means, so a reader could easily conclude that a worrying result was healthy. Although, Google removed the AI Overview for that query and for one close variant, a slightl

Follow one click. A woman taps a Facebook ad for a fertility clinic. She lands on a booking page. A tracking pixel fires the instant the page loads, logging her IP address and the exact URL, which on most hospital sites already contains the clinic's speciality in it. She scrolls to a form, picks 'IVF Consultation' from a dropdown, types her name and phone number and hits submit. In the time it took to read that paragraph, at least four different systems now hold a piece of in

A hospital's AI vendor said the tool was "fully compliant" and "healthcare-ready". That was the pitch, spoken-out loud in the sales meeting. The actual contract said something else. Customer data could be used to improve the vendor's own model. Outputs were provided "as-is", with no warranty of accuracy or bias mitigation. Vendor liability was capped at one month of fees. None of that was hidden exactly, it just wasn't what got said in the room. I'm Boudhhayan Duttaa, founder