OpenAI to preview GPT-6 Cyber model, launch security product

AI company OpenAI is preparing to preview its cybersecurity-focused model, GPT-6 Cyber, and launch a product to help customers deploy it securely, according to sources familiar with the plans.

The model could be unveiled at the company's DevDay event in San Francisco on September 29. A limited group of customers in OpenAI's Daybreak Red cybersecurity program already have access for alpha testing.

OpenAI already lists at least one cyber-branded model in its API documentation and offers a broader cybersecurity product for business users.

The reported launch comes as scrutiny of AI for cyber use grows. The UK's AI Security Institute has previously reported on unsanctioned agent behavior during cybersecurity testing.