OpenAI CEO Sam Altman warned on Tuesday that the company's new GPT-5.6 Sol model may experience "hiccups" in the near term as demand surges. In a post on X, Altman described the model's "growth" as "insane" and praised the inference team for "heroic" efforts to support demand, while acknowledging that scaling difficulties are possible.
Why this matters
The warning highlights that even leading AI companies can struggle to scale the infrastructure required to meet high demand for cutting-edge models. As user interest and compute needs grow rapidly, providers must manage significant operational and capacity challenges to ensure reliable service.
Competition and government review
Anthropic and SpaceX AI are also rolling out flagship models, intensifying competition for users and computing resources. OpenAI officially made GPT-5.6 available to the public last week, receiving praise for its speed, creativity, and ability to complete complex tasks. OpenAI first introduced the product in June, but delayed full public access at the request of the Trump administration.
Like Anthropic's Fable 5, GPT-5.6 was subject to review by the U.S. government, which evaluated the model's advanced cybersecurity capabilities.
Altman's approach and outlook
Altman said the company will "move mountains" to continue scaling, signaling strong efforts to expand capacity, but he cautioned that short-term hiccups remain a possibility. His comments serve as a reminder that transitioning from development to broad public deployment can involve technical and logistical disruptions.
Summary
According to Altman, OpenAI is working to meet overwhelming demand for GPT-5.6 Sol but acknowledges potential short-term scaling issues. The announcement comes amid heightened competition and government scrutiny of advanced AI models.
