OpenAI Launches New Model for Cybersecurity Teams

Golden Finance reported on May 8 that OpenAI announced on Thursday the release of a variant of its latest AI model, GPT-5.5-Cyber, in a limited preview for vetted cybersecurity teams. A month earlier, its competitor Anthropic garnered attention from investors and government officials with the preview version of Claude Mythos. OpenAI emphasized that the preview version of GPT-5.5-Cyber is not designed to significantly enhance cybersecurity capabilities but has been trained to be more adaptable when handling security-related tasks. The company stated that with this version tailored for the cybersecurity field, vetted teams will find it easier to leverage OpenAI's latest model for workflows such as vulnerability identification and classification, patch validation, and malware analysis. (Dongxin News Agency)