OpenAI’s rapid release of GPT-5 triggered not only a major software update but also an unexpected backlash from users mourning the loss of GPT-4o. The swift removal of the earlier model caught the community off guard and sent thousands to sign petitions demanding its return. Such an intense response reveals how digital tools with distinct “personalities” create loyalty and habits that users are reluctant to give up. Conversations online moved beyond technical improvements, touching on users’ emotional connections. The episode also highlights the challenges that companies face in anticipating and managing expectations when introducing new AI models.
Earlier announcements about OpenAI’s new models tended to generate excitement, with upgrades widely welcomed as steps toward smarter and more helpful AI. Unlike previous rollouts, though, the debut of GPT-5 was marked by public disappointment and a steep drop in confidence on prediction markets, reflecting skepticism about whether the model provided significant progress over GPT-4o. While technical bugs had troubled earlier launches, it was the immediate elimination of GPT-4o—rather than the new model’s performance alone—that fueled much of the recent controversy. This customer response contrasts with earlier patterns when new models typically led to a swift and nearly universal transition.
How Did Users React to GPT-4o Removal?
Reactions across platforms revealed that GPT-4o’s conversational tone and personality were central features for many users, sparking calls for its reinstatement. The absence of a model selector tool limited user agency, forcing reliance on GPT-5 and further heightening frustration. Petitions rapidly collected over three thousand signatures as users asked OpenAI to reconsider its decision.
Will GPT-4o Remain Available to All?
OpenAI responded by reinstating GPT-4o, but only as an option for ChatGPT Plus subscribers, who pay $20 monthly. Addressing feedback, CEO Sam Altman posted,
“Ok, we hear you all on 4o; thanks for the time to give us the feedback (and the passion!).”
The decision means the broader user base will still interact primarily with GPT-5, while those most invested in GPT-4o’s features have a paid path to access it again.
Can GPT-5 Match User Expectations for “Personality”?
OpenAI acknowledged that GPT-5’s character felt flat relative to GPT-4o, prompting a planned focus on making the model more personable. Altman shared,
“We expected some bumpiness as we roll out so many things at once. But it was a little more bumpy than we hoped for!”
He described the company’s plan to develop GPT-5’s conversational “warmth” in response to community feedback.
Recent events illustrate how AI models’ perceived personalities can shape user trust, satisfaction, and brand loyalty. For organizations developing or deploying AI, attention to not only technical ability but also user experience proves crucial. While some users welcome efficiency and intelligence upgrades, many prioritize consistency and connection—preferences that may not have been fully appreciated before. As AI models become more socially present, companies like OpenAI face the challenge of serving diverse user expectations. Those relying on AI tools for more than information—seeking guidance, companionship, or decision support—adjust slowly to sudden changes. For readers, recognizing the complexity of user attachment may guide decisions about when and how to adopt or switch AI platforms: regular backups of essential conversations, trial periods with new models, and reading user forums can help maintain agency during transitions. Watching how OpenAI and its competitors address these evolving demands could offer insight into the balance between innovation and continuity in digital tools.
- OpenAI restored GPT-4o after user petitions and critical feedback.
- GPT-5 remains default, but GPT-4o returns for paying subscribers.
- Company will work to improve GPT-5’s conversational personality.
