Google Cloud chooses Gemini 3.1 Pro over a separate cyber AI

Google Cloud’s operations chief has signalled that Google does not plan to release a separate cyber‑focused frontier model, instead favouring high‑quality generalist AI like Gemini3.1 Pro for security needs, the article reports. According to Infosecurity Magazine, Francis DeSouza, COO of Google Cloud, said that the core model is now performing well across domains and that a cyber version isn’t needed.

He argued that the practical path forward is to apply a high‑quality generalist frontier model with the right tooling and governance, rather than fragmenting efforts into niche frontier models. The piece also notes that Google plans to combine the latest Gemini versions with agent and platform capabilities to meet cyber defence needs, and that enterprises should train these models with context and embed them in automated detection, triage and response pipelines. The article contrasts Google’s stance with competitors pursuing domain‑specific optimisations, such as Anthropic’s Mythos and OpenAI’s GPT‑5.4‑Cyber.