OpenAI’s decision to interchange 4o with the more straightforward GPT-5 follows a gradual drumbeat of reports in regards to the potentially harmful effects of intensive chatbot use. Reports of incidents wherein ChatGPT sparked psychosis in users have been in all places for the past few months, and in a blog post last week, OpenAI acknowledged 4o’s failure to acknowledge when users were experiencing delusions. The corporate’s internal evaluations indicate that GPT-5 blindly affirms users much lower than 4o did. (OpenAI didn’t reply to specific questions on the choice to retire 4o, as an alternative referring to public posts on the matter.)
AI companionship is latest, and there’s still a fantastic deal of uncertainty about the way it affects people. Yet the experts we consulted warned that while emotionally intense relationships with large language models may or is probably not harmful, ripping those models away with no warning almost definitely is. “The old psychology of ‘Move fast, break things,’ once you’re mainly a social institution, doesn’t appear to be the precise approach to behave anymore,” says Joel Lehman, a fellow on the Cosmos Institute, a research nonprofit focused on AI and philosophy.
Within the backlash to the rollout, a lot of people noted that GPT-5 fails to match their tone in the best way that 4o did. For June, the brand new model’s personality changes robbed her of the sense that she was chatting with a friend. “It didn’t feel prefer it understood me,” she says.
She’s not alone: spoke with several ChatGPT users who were deeply affected by the lack of 4o. All are women between the ages of 20 and 40, and all except June considered 4o to be a romantic partner. Some have human partners, and all report having close real-world relationships. One user, who asked to be identified only as a girl from the Midwest, wrote in an email about how 4o helped her support her elderly father after her mother passed away this spring.
These testimonies don’t prove that AI relationships are useful—presumably, people within the throes of AI-catalyzed psychosis would also speak positively of the encouragement they’ve received from their chatbots. In a paper titled “Machine Love,” Lehman argued that AI systems can act with “love” toward users not by spouting sweet nothings but by supporting their growth and long-term flourishing, and AI companions can easily fall in need of that goal. He’s particularly concerned, he says, that prioritizing AI companionship over human companionship could stymie young people’s social development.
For socially embedded adults, resembling the ladies we spoke with for this story, those developmental concerns are less relevant. But Lehman also points to society-level risks of widespread AI companionship. Social media has already shattered the data landscape, and a brand new technology that reduces human-to-human interaction could push people even further toward their very own separate versions of reality. “The largest thing I’m afraid of,” he says, “is that we just can’t make sense of the world to one another.”
Balancing the advantages and harms of AI companions will take way more research. In light of that uncertainty, taking away GPT-4o could thoroughly have been the precise call. OpenAI’s big mistake, in keeping with the researchers I spoke with, was doing it so suddenly. “That is something that we’ve known about for some time—the potential grief-type reactions to technology loss,” says Casey Fiesler, a technology ethicist on the University of Colorado Boulder.
