Microsoft says the platform routes security tasks across its own, OpenAI and Anthropic models, including attacker-and-defender style vulnerability testing aimed at finding flaws before hackers do.
verifying reliability
No specialized terms available for this topic.