- OpenAI has officially disabled GPT-4o in ChatGPT
- A lot of users are feeling emotional and upset about the switch
- There’s an official #keep4o campaign underway
We knew the moment was coming, and now it’s happened: OpenAI has officially disabled the GPT-4o model inside ChatGPT, pushing all users towards the GPT-5 alternatives, and it’s hitting a lot of users hard.
A substantial chunk of ChatGPT users prefer the more emotional and warmer ChatGPT-4o experience, as it’s more suited to AI companionship and bonding. Now it’s no longer available, there’s been a widespread outpouring of sadness and anger.
“I’m grieving, like so many others for whom this model became a gateway into the world of AI,” writes one user on Reddit. Other posts talk about AI friends being “erased” and a sense of “emotional and creative collapse” without the older models.
You may like
There’s a lot of criticism for OpenAI online too, including accusations of hypocrisy – the company often claims to want to protect the mental well-being of its users, but it’s also left a lot of people feeling sad and lost this weekend.
The campaign to #keep4o
keep4o replies in the comments are scary man – by the end of this year we will be looking at an llm psychosis epidemic https://t.co/D5cFFclRcoFebruary 12, 2026
There is an ongoing campaign to keep GPT-4o available, run through Reddit and the #keep4o hashtag on social media. A Change.org petition to bring the model back has reached almost 21,000 signatures at the time of writing.
That may not seem a significant number of people compared to the millions of users who log into ChatGPT every day, but it reflects the attachment that many have formed with the ‘personality’ behind the GPT-4o model.
While to some it might seem strange to form a connection to an AI chatbot, for others it’s very real and meaningful – there are even research papers on the phenomenon, which talk about “deep socio-emotional attachments to AI systems”.
That’s something that all AI companies and society as a whole will have to grapple with in the future. It’s clear that AI bots are now capable enough to function as friends, therapists, and more – but it’s less clear just how beneficial that’s going to be for us in the long term.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.
The best business laptops for all budgets
Our top picks, based on real-world testing and comparisons
