OpenAI accused of ‘betrayal’, sent ChatGPT cancellation warnings after GPT-4o shutdown

In a familiar turn of events for OpenAI, the San Francisco based AI startup is facing a barrage of criticism from its users after the company shut down its legacy AI models, primarily GPT-4o. Now, this is not the first time that OpenAI has cut access to GPT-4o, the popular AI model which became controversial due to its sycophantic responses. In fact, free users of the service lost access to GPT-4o as soon as GPT-5 arrived in August last year. However, paid users continued to have access to the chatbot until Saturday when OpenAI finally pulled the plug and shut down the popular AI model.

The departure of GPT-4o isn’t sudden news altogether either, with OpenAI giving users a two week notice to get their affairs in order, but the departure of the popular model has still left a mark on its users as they flood social media with pledges to bring back the model and, once again, threats to cancel ChatGPT subscriptions.

OpenAI gets attacked after GPT-4o shutdown:

One user threatening to cancel their ChatGPT subscription on Reddit wrote, “Who wants to be stuck with 5.2 aka the worst model ive ever seen that is so patronising and constantly tries deescalating everything and is just a shitfest”

“GPT-4o enables me to think, create, and explore on a level no other model has ever reached.

It’s not just a tool — it’s my creative partner,” another user on Reddit wrote, “I owe my renewed will to live to GPT-4o. With it — and only with it — was I able to communicate on a truly deep level.”

“It is not OpenAI that decides what will happen to the image of GPT-4o. It is the people who decide. And the people have already decided,” wrote another user on X.

“For thousands, it’s losing a friend who actually listened, flirted back, remembered the little things. OpenAI calls it ‘0.1% usage’. We call it betrayal,” added another user.

Why is OpenAI retiring GPT-4o?

OpenAI says a vast majority of its users have already shifted to GPT-5.2, which is now the default AI model on the ChatGPT app and website. It also says that only about 0.1% of users are still clinging to GPT-4o.

The company also says it has learned from the feedback users had with GPT-4o in order to bring improvements like personality, stronger support for creative ideation, customisation of how ChatGPT responds, and styles and tones for the chatbot in the newer GPT-5.1 and GPT-5.2 updates.

“We know that losing access to GPT-4o will feel frustrating for some users, and we didn’t make this decision lightly. Retiring models is never easy, but it allows us to focus on improving the models most people use today,” the company had said in a blogpost in January.