OpenAI Launches Security Initiative with GPT-5.5-Cyber to Compete with Anthropic’s Mythos

OpenAI has launched a security initiative to compete with Anthropic’s Mythos and now offers a security review service for open-source projects. The update introduces an enhanced safety model, referred to as GPT-5.5-Cyber, though no specific technical details, release dates, or CVE IDs were provided. The initiative aims to improve security measures within OpenAI’s systems and extend support to external open-source projects. No explicit impacts, affected systems, or vulnerabilities were described in the article. The announcement was reported by heise.de without additional numerical or technical data.