AI Proxy

A design pattern in which an AI system acts on a user's behalf within a social, communicative, or interpretive setting — for example, generating a facial expression, voice, or written reply that represents the user to others — rather than merely assisting the user with a personal task. In accessibility contexts, AI proxies can enable disabled users to participate in interaction channels they would otherwise be locked out of (e.g., visual facial expression for blind users, spoken voice for users who do not use speech), but they introduce new questions about consent, personalisation, misrepresentation, and verification: the user may not be able to confirm what the proxy is saying on their behalf. Good AI-proxy design therefore typically requires upfront user-defined preferences, ongoing transparency, and the ability to override or pause.