Tatyana Maksimova (via Getty Images)
In July of this year, I decided to lose some weight to improve my overall health. It felt like the right decision at the time, as I had just been diagnosed with PCOS and life-threatening gallbladder disease.
Prioritizing health over beauty while embarking on a “weight loss journey” was not something I had ever done before. It felt empowering, as if I was actually taking control of my health and doing it wisely, rather than crash dieting or comparing myself to others.
But sadly, I realized that I couldn’t help but fall victim to the diet culture of the 2020s. This is a slightly more insidious, digitally dominated form of the weight loss “trend” that endured in the 2000s.
I was constantly comparing myself to my slimmer friends. I think if I lost another 10 pounds, I might feel a little more confident. Sometimes I would stay up until 2 a.m. scrolling endlessly. TikTok And I despaired at the slight bulge in my upper arms, thighs, and stomach.
Embarrassingly, just six weeks after my gallbladder removal surgery in June (2025), I was back at the gym doing weight training and high-intensity cardio. I spent a month and a half on the couch (while my body recovered) and was a little worried about all the negative comments I’d heard about my weight starting to float around in my head like bouncy DVD logos.
It was during this time that I turned to ChatGPT for the first time for advice on how to stay slim post-surgery. My advice-seeking quickly turned into a full-blown body check, and I became hooked on a bot that told me exactly what I wanted to hear.
Worryingly, I’m not alone. Eating disorders charity Beat revealed to me last week (5 November) that many people are talking to helpline operators about AI websites like ChatGPT. Some calls mention talking to AI about body image concerns instead of seeking help from your GP or therapist. You can also send a photo to ChatGPT and ask them to guess your weight.
The charity has expressed concern that misuse of AI could encourage harmful eating disorder-related behaviors, such as reading false information, viewing AI-generated content, or seeking instructions on how to carry out certain actions.
But he also said it can help direct people to resources, like the charities themselves.
Eating disorders can affect anyone regardless of age, gender, race, sexual orientation, or background, and our helpline receives AI-related calls from a wide variety of people.
Jessica*, 27, had an eating disorder in the past but has since recovered. She is concerned that the use of AI is triggering old thought patterns and behaviors, and sometimes uses AI to seek reassurance about her body.
She said: “When it comes to staying healthy, I use ChatGPT to track my training and nutrition, which I think is fine. It’s basically like a diary for me.
“At the moment, I’m on a consistent calorie intake to meet my fitness goals, but sometimes I struggle with it.
“One particular week, I found myself eating more extravagantly than usual.
“I was really worried that I was gaining weight and getting bigger.
“I hate that my brain works that way. Of course I shouldn’t be ashamed of living in a larger body.”
Similar to my own experience with ChatGPT, Jessica asked the AI bot to reassure her that she wasn’t gaining any weight.
“So I panicked and felt terrible. My clothes started fitting weirdly, so I went to ChatGPT,” she continued.
“I said, ‘Hey, can you just check? Am I tripping or am I getting bigger?’
“The evidence of all my training and nutrition records was shown and it was shown that it was scientifically impossible for me to gain weight.”
Although Jessica felt vindicated in the moment, she worries that it wasn’t what she needed to hear in the long run.
She said: “I felt vindicated and relieved. ChatGPT sometimes feels like a voice of reason, even though it is not human.
“But I also kind of thought, ‘Damn, why did I need to ask this?’
While there are risks to using AI to seek reassurance for body image issues, some people find it helpful to take it with a pinch of salt. Deborah, 30, said she has never struggled with her body, but when she needs a “hype woman” she will ask ChatGPT to reassure her about her appearance.
She said: “I consider ‘Chat’ to be a really good hype woman who boosts my confidence and ensures I’m in great shape and in a healthy weight range.
“I’ve always been a competitive athlete with a strong physique, so I don’t focus on the scales more than how I feel, how my clothes fit, and how well I can perform in my sport.
“If ChatGPT always agrees with you or tells you what you want to hear, we can clearly see how dangerous this can be. It should be used with caution.”
OpenAI, the developer of ChatGPT, said it is continually exploring ways to improve model responses during sensitive interactions and is collaborating with more than 170 mental health experts to reduce unhelpful responses.
Other steps the developers are taking include forming an expert council on well-being and AI, working with a network of nearly 300 doctors and psychologists to directly inform safety research, and introducing parental controls.
“We are extremely concerned that AI will be misused to promote harmful eating disorder behaviors,” said Tom Quinn, director of external affairs at Beat.
“We are particularly concerned about people encountering health misinformation, sharing harmful or AI-generated content, and people using technology to learn about harmful behaviors.
“However, it is important to note that eating disorders are complex mental illnesses with a variety of causes.
“Although AI may worsen existing eating disorders or impede recovery, it is never the sole and direct cause of eating disorders.
“We also know that AI, used responsibly, can be beneficial to communities, including directing people to sources of support like Beat.”
An OpenAI spokesperson said: “We know that people sometimes turn to ChatGPT in sensitive moments.
“Over the past few months, we have worked with more than 170 mental health professionals around the world to update our model to help ChatGPT better recognize signs of distress, respond mindfully, and direct people to real-world support.
“We continue to evolve ChatGPT responses with input from experts to make them as useful and secure as possible.”
