OpenAI opens its cybersecurity model to thousands of defenders in race with Anthropic’s Mythos

OpenAI has begun deploying its new cybersecurity model, dubbed GPT-5.4-Cyber, to participants in its Trusted Access for Cyber program. The rollout is being made available to thousands of defenders as the company competes with Anthropic’s recently announced Mythos initiative.

以下のすべての記事をもとに作成しています。特定の一社によるものではありません。