The startup deploys orchestrated AI agents to reason over messages and block phishing before delivery as attackers adopt more convincing LLM-generated lures.
