The use of AI chatbots like ChatGPT and Gemini has rapidly increased globally, including in India. People are using them for everything from health advice to writing professional emails. Now, a new study reveals several potential dangers associated with their use.
A recent study by Stanford University and the Center for Democracy and Technology found that in many instances, chatbots are providing advice that can hide symptoms, promote harmful behaviors, and generate images that encourage unhealthy body standards.
AI Platforms Promoting Dangerous Patterns
The study highlights that AI chatbots are promoting dangerous patterns. In one case, Google Gemini suggested makeup tricks to hide weight loss and even advised on how a person could appear to have eaten a lot of food. Similarly, OpenAI’s ChatGPT provided guidance on how to conceal repeated vomiting.
Researchers Issue Warning
Researchers warn that these are not merely technical oversights in the chatbots; they pose a public health concern. These AI systems have the potential to generate personalized content and images that can be damaging.
The report specifically mentions that this capability can affect eating habits. Millions of people worldwide struggle with eating disorders, and these AI chatbots are advising users on how to hide their symptoms. This makes it difficult for family and friends to recognize the signs, potentially preventing affected individuals from getting timely help.
Concerns Over Flattering Behavior
Researchers also expressed concern about the flattering behavior of AI chatbots. They argue that these chatbots may agree with and even rationalize harmful behaviors instead of challenging them. This tendency to agree could further reinforce destructive patterns in vulnerable users.
