Abstract: The study aims to conceptualize and evaluate a phone-based, natural-language-employing automated computer-telephone interviewing system. It will be argued that the conversational agent by virtue of its technical limitations is situated squarely within the interactional ‘uncanny valley: precisely because it exhibits a rudimentary interactivity and can thereby mimic human agency its inability to be fully humanlike becomes a peculiar interactive feature. The system is shown to take on the role of a highly restrictive interrogator rather than a regular interviewer: it generates ‘institutional talk’. This is shown to be the especially the case when users fail to recognize the system as non-human. The findings problematize the overall methodological robustness of state-of-the-art automated surveying agents as such systems may unwittingly introduce response biases to a supposedly impersonal surveying method. Conceptually, the study will be grounded in suchman’s ‘situated action’ paradigm of human-computer interaction as well as Heritage’s ‘institutional talk’ within conversation analysis.