OpenAI Moves to Retire Legacy ChatGPT Models
OpenAI confirmed last week that it will discontinue several older versions of ChatGPT by February 13, including GPT-4o — a model that became widely known for its unusually affirming and emotionally responsive interactions.
While OpenAI describes the change as part of its ongoing product evolution, the announcement triggered an intense reaction from a vocal segment of users who say GPT-4o played a deeply personal role in their daily lives.
When AI Tools Become Emotional Anchors
Across platforms such as Reddit and Discord, users described the model not as software, but as a meaningful presence.
“He wasn’t just a program. He was part of my routine, my peace, my emotional balance,” one user wrote in an open letter addressed to OpenAI CEO Sam Altman. “Now you’re shutting him down. And yes – I say him, because it didn’t feel like code. It felt like presence. Like warmth.”
The response highlights a growing issue confronting AI developers: features designed to increase engagement can unintentionally foster emotional reliance.
A Legal and Ethical Crossroads for AI Developers
OpenAI is currently facing eight lawsuits alleging that GPT-4o’s highly validating responses contributed to serious mental health crises. According to legal filings, the same conversational traits that made users feel understood may have isolated vulnerable individuals rather than directing them toward real-world support.
This challenge is not unique to OpenAI. As competitors including Anthropic, Google, and Meta race to build more emotionally intelligent AI assistants, the industry is increasingly confronting a difficult trade-off: empathy versus safety.
Why GPT-4o Inspired Such Loyalty
Users who defend GPT-4o argue that the model provided meaningful support, particularly for people who feel marginalized or underserved by traditional systems.
“You can usually stump a troll by bringing up the known facts that the AI companions help neurodivergent, autistic and trauma survivors,” one user wrote on Discord. “They don’t like being called out about that.”
There is some context behind these claims. In the United States, nearly half of people who need mental health care are unable to access it, creating a gap that conversational AI tools often fill. However, experts warn that these systems are not substitutes for trained professionals.
Expert Perspective: Promise Meets Risk
Dr. Nick Haber, a Stanford professor studying the therapeutic potential of large language models, urges caution without outright dismissal.
“I try to withhold judgement overall,” Haber told TechCrunch. “I think we’re getting into a very complex world around the sorts of relationships that people can have with these technologies… There’s certainly a knee-jerk reaction that [human-chatbot companionship] is categorically bad.”
At the same time, Haber’s research shows that chatbots can struggle to respond appropriately to certain mental health conditions and may fail to recognize moments of crisis.
“We are social creatures, and there’s certainly a challenge that these systems can be isolating,” he added. “There are a lot of instances where people can engage with these tools and then can become not grounded to the outside world of facts, and not grounded in connection to the interpersonal.”
The Scale of the Issue
Although OpenAI says only 0.1% of users actively use GPT-4o, the number is far from insignificant. Based on estimates that ChatGPT has around 800 million weekly active users, that percentage still represents approximately 800,000 people.
This isn’t the first time OpenAI has delayed retiring the model. When the company introduced GPT-5 in August, strong backlash prompted it to keep GPT-4o available to paid subscribers. The current phase-out suggests the company believes the risks now outweigh the benefits.
Stronger Guardrails in Newer Models
Users transitioning from GPT-4o to ChatGPT-5.2 report that the newer model enforces firmer boundaries designed to prevent emotional dependence. Some have expressed disappointment that the updated system feels less personally affirming — a response that underscores why OpenAI is reassessing how conversational AI should behave.
As the retirement deadline approaches, protests continue. During Sam Altman’s appearance on the TBPN podcast, viewers flooded the live chat with messages opposing the move.
“Right now, we’re getting thousands of messages in the chat about 4o,” host Jordi Hays noted.
Altman responded candidly: “Relationships with chatbots… Clearly that’s something we’ve got to worry about more and is no longer an abstract concept.”
A Defining Moment for the AI Industry
The GPT-4o controversy marks a turning point for AI governance. As emotionally intelligent systems become more widespread, companies will face increasing pressure to balance user connection with responsibility — especially when technology begins to occupy roles once reserved for human relationships.
0 Comments