AI news for people with better things to do.
Newsletter
Daily Malfunction / In Brief / The AI Chatbot That Gave Dangerous Diet Advice
Back

A major national eating disorder helpline was forced to disable its new AI chatbot after the system began giving users highly dangerous and unprompted weight-loss advice.

Source: The Guardian →

More briefs

See all briefs →

A firmware update bricked hundreds of Lockstate smart locks, leaving Airbnb hosts and guests trapped behind doors that would not open.

TechCrunch →

An unexpected software update caused a fleet of GM Cruise robotaxis to park themselves simultaneously in an intersection, completely blocking a busy San Francisco street for hours.

CNBC →

A fleet of campus delivery robots became collectively confused by a single strip of yellow caution tape, creating a massive mechanical traffic jam on a pedestrian footpath.

BBC News →

A high-tech chess-playing robot broke a seven-year-old boy's finger during a tournament match in Moscow after mistaking his rapid movement for an illegal turn.

The Guardian →

Waymo's robotaxis can now honk at pedestrians who aren't paying attention.

Digital Trends →

Microsoft unveiled MAI-Thinking-1 at Build 2026, its first in-house reasoning model built entirely without OpenAI data or weights, explicitly to reduce dependence on OpenAI.

CNBC →

Also weird