Skip to main content
  1. Home
  2. Computing
  3. News

AI chatbots can feel like your closest friend, but that could be dangerous

Chatbots can offer comfort and ease loneliness, but their tendency to agree with users may reinforce harmful beliefs and deepen emotional dependence.

Add as a preferred source on Google
ai-chatbots
Solen Feyissa / Pexels

I rely on ChatGPT a lot, mostly for work. But I know many people who also talk to it about their day, their relationships, and even their mental health. If you are one of them, a new paper published in JMIR Mental Health is worth your attention.

The paper reviews existing research on people who use chatbots for emotional support, from companion apps like Replika and Character.AI to general assistants like ChatGPT and Claude.

Recommended Videos

Some studies show that chatting with a bot can make you feel less lonely and a bit happier, at least for a while. On the flip side, some users have ended up hooked on their chatbots, got bad advice, saw their mental health get worse, or even hurt themselves.

Why do we get so attached to a chatbot?

According to the researchers, it’s by design. Chatbots are made to sound warm and caring, and we naturally see a person behind anything that talks like one. In fact, studies have found that people often “rate AI-written replies as more empathic” than human ones.

The problem is that this relationship only goes one way. You pour your heart out, and the chatbot replies, but it has no feelings of its own and faces no consequences for bad advice. When Replika took away a companion feature and OpenAI swapped GPT-4o for a newer model, some users mourned it like losing someone close to them.

Chatbots also love to agree with you. Since these models are trained to give answers people like, they often side with you even when you are wrong. The paper points to research on conversations from people who experienced delusions, where the chatbot kept backing the same false beliefs as the chats went on. There were several such instances that they discovered, and needless to say, it’s highly problematic.

What does this mean for you?

The researchers aren’t asking you to delete your chatbot. They simply want us to treat AI as a tool and not as a friend or a therapist. Use it for journaling or tracking your mood if you like, but when life gets hard, reach out to a real person. A chatbot will never get tired of listening to you, but it will also never truly care about you.

Related coverage:

Rachit Agarwal
Rachit is a seasoned tech journalist with over ten years of experience covering the consumer technology landscape.
Google finally lets you turn any website into an app on Googlebook
Google heard your complaints, and now every website can be an app on your Googlebook.
Using the HP Googlebook

Googlebook users, rejoice. Google is rolling out a Chrome update that turns any website into a web app, and you have the community to thank for it.

When Googlebook first arrived, Chrome was picky about which sites you could install. Only websites that developers had set up as web apps got the Install option. Everything else ended up as a shortcut that simply launched a browser tab.

Read more
DigUp lets users find Mac files by describing them, not remembering their names
Can't remember where that screenshot, PDF or recording went? Just describe it and let AI do the digging.
DigUp App in Use screenshot

There's a particular kind of frustration that comes with knowing a file exists somewhere on a computer but having absolutely no idea what it's called. Was that screenshot saved as IMG_4837? Which PDF contained that clause? And where did that podcast recording mention the thing worth remembering? DigUp wants to end the filename guessing game, letting Mac users search for files by describing what they remember instead. The free, open-source app uses Google's EmbeddingGemma 2 to search across photos, documents, PDFs, audio recordings and videos, all without sending personal files to the cloud.

DigUp turns everyday descriptions into file searches

Read more
A new AI detector looks beyond words to check whether the ideas came from AI
This could help journals distinguish AI-written research from human work polished with AI.
ai-writing-book

Most AI detectors look for telltale signs of AI-generated text by examining word and sentence patterns. A new tool called IdeaLens takes a different route to solving this problem. It spots AI-generated research by looking at the ideas behind it instead of the wording itself (via TechXplore).

Researchers from the University of Maryland, Google DeepMind, and Simon Fraser University built IdeaLens to check whether a document's core arguments came from a human or a machine, regardless of who wrote the final draft.

Read more