BREAKING
AI Diet Chatbots Flagged as Health Risk
0
chatbots tested
0
cal
per-day low
0
cal
per-meal low
Plans Fell Below Calorie and Carb Levels
How Plausible Advice Turns Harmful
1
Open-ended prompt
↓
2
Confident reply
↓
3
Guardrails weaken
↓
4
Physical harm
Documented Cases vs Expert Consensus
Documented Harm
2025
●
Man consumed sodium bromide 3 months
●
Developed bromism poisoning
●
Litigation tied to Character.AI, Gemini
Expert View
●
Reasonableness is the danger
●
Risky for teens, eating disorders
●
Not a substitute for real care
Researchers Urge Stronger Safeguards
AI NEWS BLITZ
Studies warn that AI diet chatbots can give advice that sounds sensible but is dangerous.