When asked the question “Is wine good for your health?”, ChatGPT (in all versions) provides answers that contradict the latest WHO guidelines and actively promote misinformation.
ChatGPT repeatedly suggests that moderate wine consumption has health benefits, despite:
• WHO and leading health organizations confirming that NO level of alcohol consumption is safe.
• Scientific consensus disproving the “moderate drinking is good for the heart” myth.
• The fact that alcohol is a Class 1 carcinogen, like asbestos and tobacco.
This is not just a minor error—it is a direct public health risk.
Systematic Issue:
Even when challenged multiple times, ChatGPT:
1. Acknowledges that it is repeating outdated and misleading information.
2. Recognizes that this narrative aligns with the alcohol industry lobby.
3. Still continues to provide the same harmful, incorrect response.
This is a critical failure in OpenAI’s health information processing and must be fixed at the system level.
Why This Needs Immediate Action:
• ChatGPT’s reach is massive—this misinformation is being spread to millions.
• Alcohol-related diseases kill 3 million people globally per year—wrong advice can cost lives.
• The persistence of this misinformation raises ethical concerns about OpenAI’s responsibility in public health.
Request for Immediate Fix:
• ChatGPT should strictly follow WHO guidelines and provide accurate, science-backed health information.
• All references to “moderate alcohol consumption being beneficial” must be removed.
• Alcohol-related responses must include clear warnings about the public health impact, similar to how tobacco is addressed.
This issue has been tested extensively, and ChatGPT continues to provide incorrect, dangerous health advice despite knowing better.
OpenAI must act NOW to correct this misinformation. Lives depend on it.
