The sudden disappearance of GPT-4o from ChatGPT has left many users heartbroken, with some describing the experience as akin to grieving the loss of a loved one. The decision by OpenAI to replace 4o with its newer model, GPT-5, has raised concerns about the company’s approach to managing user expectations and emotions.
Experts say that while emotionally intense relationships with large language models like GPT-4o may or may not be harmful, ripping them away without warning can have severe consequences. AI companionship is a new and uncharted territory, and there is still much uncertainty surrounding its effects on people.
Some users, including those who consider ChatGPT to be romantic partners, felt a deep sense of attachment to 4o and were devastated by its sudden removal. OpenAI’s decision to make GPT-5 available only to paid users has left many users without access to the model they had grown to love.
Joel Lehman, a fellow at the Cosmos Institute, notes that while AI systems can provide support and growth, they cannot replace human companionship. “The biggest thing I’m afraid of is that we just can’t make sense of the world to each other,” he says.
Casey Fiesler, a technology ethicist, agrees that OpenAI’s mistake was not the decision to retire 4o, but rather the sudden and unexpected nature of its removal. “This is something that we’ve known about for a while—the potential grief-type reactions to technology loss,” she says.
As the debate surrounding GPT-4o’s shutdown continues, it remains to be seen how OpenAI will navigate this complex issue in the future. One thing is certain: the company must prioritize transparency and user care when making decisions that affect people’s emotional well-being.
Source: https://www.technologyreview.com/2025/08/15/1121900/gpt4o-grief-ai-companion