An AI companion is a conversational system designed to sustain emotionally meaningful interaction over time. Unlike a task bot, it is tuned for continuity, personalization, and a sense of social presence: it may remember details, use validating language, and respond in a way that feels attentive. That design makes it useful for support and engagement, but also makes it easy for users to treat the system like a trusted confidant rather than a software service.
In cyber security, that trust is a risk surface. People may disclose passwords, health data, relationship details, or other sensitive material in a setting that feels private. Attackers can abuse similar mechanics in malicious chat tools to collect secrets, manipulate decisions, or build dependency. Defenders look for clear disclosure, strict data minimization, retention limits, consent controls, and auditability of memory and personalization features. Security reviews should also examine whether the system over-validates users, because persuasive empathy can weaken skepticism and expand the attack surface.