OpenAI releases new language model GPT-5.6 Sol after security review

Tech firm OpenAI has launched its latest language model, GPT-5.6 Sol, making it available for public use after a month-long delay. The release was initially postponed at the request of the U.S. government, which had expressed concerns regarding national security implications and the potential for misuse of the advanced AI technology.

Following additional security assessments and testing, the U.S. government granted approval for the broader rollout of the new model. OpenAI stated that GPT-5.6 Sol represents their most sophisticated model to date, particularly in the realm of cybersecurity.

In addition to GPT-5.6 Sol, OpenAI is also releasing two other versions, Terra and Luna. These models are specifically designed to assist with everyday work tasks, broadening the application of their AI technology beyond specialized fields.

Rasmus Larsen, a senior AI specialist at the Alexandra Institute, commented on the new model's capabilities. He suggested that GPT-5.6 Sol appears to offer a significant advancement over OpenAI's previous models. "It seems like a relatively big leap if you are to believe their benchmark," Larsen noted, referring to performance evaluations used to assess language model responses to prompts.