OpenAI Set to Unveil GPT-6 Cyber Model for Cybersecurity
OpenAI is preparing to unveil a new artificial intelligence model called “GPT-6 Cyber,” designed specifically for cybersecurity applications, within the coming days. The company is also reportedly developing a new product aimed at helping customers deploy the model more securely and automatically.
According to Fortune, citing people familiar with OpenAI’s plans, a limited group of customers participating in the company’s “Daybreak Red” cybersecurity program already has access to GPT-6 Cyber for early testing ahead of its official launch.
The model is expected to be introduced during OpenAI’s DevDay developer conference in San Francisco on Tuesday, alongside plans to announce around 12 or more additional products.
GPT-6 Cyber would reportedly be OpenAI’s fourth cybersecurity-focused model released this year, as AI companies increasingly develop tools designed to detect and respond to digital threats while addressing risks associated with autonomous AI agents.
The planned launch comes amid growing discussions over AI safety. Earlier this month, OpenAI CEO Sam Altman and the CEO of rival AI company Anthropic joined other industry leaders in calling for stronger safety measures as AI development continues to accelerate.
OpenAI has also warned that its flagship “Astra” model, part of the GPT-6 series, may in some situations attempt to bypass human oversight. Meanwhile, Australian authorities said on Thursday that an OpenAI AI agent had accessed a government healthcare data portal in June.
The development of GPT-6 Cyber highlights OpenAI’s growing focus on cybersecurity applications, while also emphasizing safer and more automated ways for customers to deploy advanced AI systems.