OpenAI Grants Early Access to Hacking Model GPT-5.6-Cyber for Select Partners

OpenAI has given a select group of partners early access to a specialized AI model called GPT-5.6-Cyber, which is designed for hacking. The move is part of OpenAI's efforts to enhance cybersecurity, allowing chosen organizations to test the model's capabilities in identifying vulnerabilities and strengthening defenses. The model is expected to assist in proactive security measures, although details on the specific partners and the exact nature of the access have not been disclosed by OpenAI.