‘This was on us’: Sam Altman on what went wrong with GPT-5 launch and why users are not liking it

OpenAI CEO Sam Altman has admitted that the launch of GPT-5, the company’s most advanced AI model to date, did not unfold as planned. Speaking to reporters, Altman acknowledged the wave of criticism that followed the rollout and explained what the company is doing to regain user trust.

The main point of contention came from OpenAI’s decision to discontinue all older models at once, forcing users onto GPT-5. Many complained that the new system delivered shorter, less engaging answers compared to GPT-4o. Within days, OpenAI reversed course, reinstating GPT-4o for ChatGPT Plus subscribers while maintaining GPT-5’s “Standard” and “Thinking” versions.

Altman admitted the move was poorly judged. “I think we totally screwed up some things on the rollout,” he said, adding that removing older models had been “the wrong decision.”

Social media platforms quickly filled with negative feedback, with long-time ChatGPT users expressing frustration at the sudden shift. The model picker—previously removed as part of the GPT-5 transition—was restored after significant pushback. OpenAI also promised updates to make GPT-5 responses “warmer” and more aligned with user expectations.

Altman acknowledged the scale of the challenge, noting, “We’ve learned a lesson about what it means to upgrade a product for hundreds of millions of people in one day, and the differences in the kinds of attachment people have with this product versus previous products.”

Despite the criticism, Altman pointed out that usage surged immediately after launch. “ChatGPT API traffic doubled in the 48 hours following the new model launch, and app usage hit a complete high,” he said.

OpenAI recently reported reaching 700 million weekly active users, and Altman stressed that the company remains committed to refining GPT-5. He suggested the backlash is part of a broader learning curve for products that people now rely on daily.

While OpenAI adjusts its approach, Altman emphasized that innovation will continue. “We’re listening carefully, and we’ll make the model better,” he said, underscoring that GPT-5’s rollout is far from the company’s final word on the future of AI assistants.