OpenAI delays launch of its first open model yet again, Sam Altman explains why

In a recent announcement, OpenAI's CEO Sam Altman revealed that the launch of the company's highly anticipated open-weight model has been postponed once again. Initially slated for release next week, Altman explained that the delay is necessary to conduct further safety evaluations and to address high-risk areas associated with the model. In a message shared on X, Altman expressed, "We planned to launch our open-weight model next week. However, we are delaying it to allow more time for additional safety tests and to review high-risk areas. We aren’t sure how long this will take." He emphasized the importance of getting this right, noting, "Once weights are out, they can’t be pulled back. This is new for us, and we want to ensure it meets the community's expectations. I apologize for this disappointing news; we are working tirelessly!" This latest delay follows a previous announcement last month in which Altman mentioned that an unexpected breakthrough from the research team warranted additional time. He had shared, "Our research team did something unexpected and quite amazing, and we believe it will be worth the wait, but it needs a bit longer." The pressure on OpenAI to deliver an open-weight model has intensified, particularly after the success of DeepSeek’s open-source AI models earlier this year. OpenAI's original mission focused on open-source principles and collaboration, but with competitors like Alibaba's Qwen and Meta's Llama offering customizable open-weight models, the urgency has increased. Moreover, enterprise users are leaning towards open-weight models for their enhanced flexibility and control. In response to criticisms, Altman previously acknowledged that OpenAI had been on the 'wrong side of history' regarding the open-source debate, and he reiterated the company’s commitment to developing a model with reasoning capabilities. He noted that while building an open-weight model has been a priority, other projects had taken precedence. As OpenAI continues to navigate these challenges, the tech community remains eager for updates on the future of open-weight models.