OpenAI releases security‑focused model

OpenAI has launched a cyber‑specific variant of GPT‑5.4 aimed at security operations, with access restricted to users who verify they are cybersecurity defenders. The product positions itself as a gated, domain‑specific model for defensive workflows. (siliconrepublic.com)