OpenAI announces cybersecurity model gpt-6 cyber for preview

OpenAI is set to preview its cybersecurity-focused model, GPT-6 Cyber, within days, according to multiple sources. The model, along with a complementary product, will be unveiled at the OpenAI Developer Conference in San Francisco on September 29, potentially earlier. GPT-6 Cyber will be the fourth cybersecurity model released by OpenAI in 2026. A few customers in the Daybreak Red program have already gained access to the model for alpha testing. Daybreak Red provides access to OpenAI's most advanced cybersecurity tools, while Daybreak Blue focuses on general protection scenarios. OpenAI also plans to launch over a dozen other products at the conference, primarily targeting enterprise and consumer markets. The model is expected to be officially launched within months. The company's cybersecurity initiatives continue to expand, with GPT-6 Cyber marking a significant step in its efforts to enhance digital security through AI.