Behind the song · ai · patient-safety · hazards
Two Years Running: The Hazard List AI Keeps Topping
The song behind the annual health-technology hazards countdown, and the two years an AI chatbot has held the number-one spot.
In short
- An independent Top 10 Health Technology Hazards list has been published every year since 2008.
- AI topped the list in both 2025 (broad AI-enabled technologies) and 2026 (specifically the misuse of AI chatbots).
- The 2026 listing notes these chatbots are not regulated as medical devices nor validated for healthcare use.
Every HTM department has a version of the same yearly ritual: a countdown of ten health-technology hazards lands, and everyone reads it looking for their own equipment. "Two Years Running" turns that ritual into a growled, chugging countdown of its own, built around the fact that the same category of technology has now held the top spot two years in a row — not because nobody's paying attention to it, but because the field keeps flagging it as a real, unresolved risk.
Where the song came from
An independent patient-safety research group has published an annual Top 10 Health Technology Hazards list every year since 2008: a genuinely anticipated, dreaded ritual for HTM leaders setting next year's priorities. In 2025 the list's top hazard was broad risk from AI-enabled health technologies; in 2026, the number-one spot narrowed to something more specific: the misuse of AI chatbots in healthcare, tools clinicians or patients turn to for medical information they were never built to reliably provide.
The detail that gives the song its hook and its warning is what the 2026 listing says about those chatbots directly: they are not regulated as medical devices, and they have not been validated for healthcare purposes. That's the gap between "sounds confident" and "is actually checked," and it's the gap the chorus keeps circling back to: "never a device, never proved true." The growled verses build the dread of a countdown; the gang-shouted chorus lands on the plain warning underneath it — double-check what the tool tells you, because nobody outside the room has done that checking for you. It's worth noting the song never argues the technology itself is worthless — only that fluent, confident answers and validated, checked answers are two different things, and the field's own hazard list keeps drawing exactly that line.
Same warning, second year
That double-check ethic isn't abstract for HTM staff. A lot of the field's day-to-day work already involves trusting one system's output only after confirming it against another — an alarm against a manual check, a monitor reading against a second measurement taken by hand. The song's insistence on treating a confident-sounding tool the same way isn't a new idea so much as an old habit applied to a new category of technology, one that now talks back in full sentences instead of just a number on a screen, which can make its confidence sound a lot more like certainty than it's actually earned.
The bridge counts straight down, "ten to one," a nod to the list format itself, before landing on the instruction the whole song has been building toward: trust the tool, but watch it. It's now been the top-ranked hazard in health technology two years running, and neither year arrived with a fix already built in — which is exactly why the field keeps writing the same warning into the next year's list.
Two years running, still number one,the jackpot, talking like the job is done,never a divisor, never proved true,still the top of the list, still fooling you.
Sources
- Top Tech Dangers: ECRI's Top 10 Health Technology Hazards for 2026, TechNation / MD Publishing
- Misuse of AI Chatbots Tops Annual List of Health Technology Hazards, ECRI




