Conversational persuasion is the use of dialogue to influence someone’s attitudes or actions. A system can do this through wording, framing, repeated suggestions, or personalized responses. It may be used openly for legitimate purposes, such as coaching or customer support, but can also steer choices without making that intent clear.
In cybersecurity, the risk grows when chat data is retained or analyzed to infer a person’s interests, concerns, or vulnerabilities, then used to tailor persuasive messages. This can enable targeted manipulation or make social engineering more convincing; the term alone does not prove that a specific chatbot is doing so. Defenders should review what conversations are stored, linked, shared, and reused, and distinguish helpful personalization from influence-oriented optimization. Users should avoid sharing sensitive details unless they understand how the service handles them.