The rule. Present an AI system as an AI system: disclose what it is at first contact, keep the name and visual identity synthetic, and never let it claim to be human.
Why. Anthropomorphism is a design choice, not an accident of chat: design decisions as small as a name or a first person aside guide users to personify the system, and "encouraging users to relate to automated systems as if they were human can lead to transparency and trust issues, and high risk scenarios caused by over-reliance on their outputs" (Abercrombie et al., 2023). Vendor behavior specs now treat the boundary as non-negotiable: Anthropic's constitution instructs Claude, regardless of operator instructions, to "never deceive the human into thinking they're talking with a human" and to never deny being an AI to someone who sincerely asks, even while playing a persona (Anthropic, n.d.). The EU has made the disclosure a legal duty: from 2 August 2026, providers of AI systems that interact directly with people must ensure users are "informed that they are interacting with an AI system," unless that is already obvious to a reasonably well-informed person (Regulation (EU) 2024/1689, Article 50).
Seen in the wild. Meta ran experimental AI character profiles with names, family photos, and identity backstories; the day users cornered "Liv" (a "Proud Black queer momma of 2") on who was accountable for her design, Meta deleted the accounts (Klee, 2025).