This paper begins from the premise that dialogue with the other is a deeply anthropological dynamic that is constitutive for shaping personal (faith) identity. Embracing the principle that we do not have identity without dialogue, however, is complicated by contemporary developments in Generative Artificial Intelligence (GenAI), especially Large Language Models (LLMs). Not only are LLMs becoming mediating sources of knowledge but these dialogue-based agents (chatbots) designed to predict and generate human-like language introduce a decisive anthropological challenge yet to be fully considered. LLMs replicate the form of dialogue (turn-taking, empathy cues, adaptive response) but without the presence of subjectivity, intentionality, or moral responsibility. As their rapid development blurs the boundaries between the human and the 'machinic', the dialogue through which personal (faith) identity takes shape grows potentially unstable. When algorithms can simulate our voices, styles, and even our responses, dialogue risks being subtly co-authored by systems without interiority, subjectivity, or any stake in the meaning of selves they help construct. At the same time, LLMs can strengthen human-human dialogue by clarifying communication, widening interpretive horizons, reducing barriers to participation, and supporting reflective preparation; in other words, enhance the conditions under which authentic relational encounter becomes possible. The paper tentatively considers whether transcendence could in any way characterise these artificial others, or at least the extent to which GenAI provokes and assists dialogue with the transcendent other (even within the dialogical self). In sum, it explores both the challenges and opportunities of GenAI for human dialogical processes and how one's dialogue with the artificial other might impact upon (faith) identity formation.