Internet of Shit: AI Poop Analysis App Offered to Sell Me Database of Its Users’ Poops

What happened

An AI-powered app that analyzes stool images to offer health insights was found offering access to its entire database of user-uploaded toilet photos for sale. This database reportedly contained 150,000 images of human stools, collected to train and improve the app’s diagnostic models. The discovery came when the app’s operator revealed hoarding this unusually large and sensitive dataset, then attempted to monetize it beyond the app’s original scope.

Why it matters

This incident exposes a major privacy and security risk in health-focused AI startups. Stool images might sound trivial or funny, but they are deeply personal medical data. Selling access to such data raises serious questions about user consent, data protection, and how health AI companies manage sensitive imagery. For anyone building or investing in consumer health AI, it shows how lucrative health data can tempt operators to take questionable shortcuts. It also forces tighter scrutiny on compliance with privacy laws like HIPAA or GDPR, which regulate sharing medical information.

What to watch next

Expect regulators and privacy watchdogs to pay closer attention to AI health apps that collect unconventional personal data. Builders should anticipate demands for stronger data governance and transparency around how training datasets are sourced, stored, and shared. Investors and customers need to factor in heightened risks around data leaks or unauthorized sales. The line between helpful health AI and invasive data harvesting remains blurry, and this case makes clear that responsible data handling can no longer be an afterthought.

AI Quick Briefs Editorial Desk