AI Disclosure

You are talking to software

Every response in a Talk With Jesus conversation — the voice, the words, the pauses — is generated in real time by a large language model hosted through ElevenLabs Agents and spoken by a synthetic ElevenLabs voice, guided by written instructions we authored. The animated figure is a designed illustration whose mouth movements are computed from the audio. Nothing in this product is Jesus, a channel to Jesus, a divine message, or a supernatural event.

Why we made it anyway

The same way a devotional book or a hymn can help a person pray without being God, a carefully shaped conversation can give people words, calm, and a nudge toward scripture and community. That is the whole ambition. We chose the second person — a companion speaking in the manner of the Gospels — because many people find it easier to be honest aloud with a “you” than into a void. That is a design choice about human psychology, not a theological claim.

Its real limits

The model can be wrong — about facts, about scripture, about you. It does not know you beyond the current session, holds no memory of past conversations (we store no transcripts), cannot perceive your situation, and has no authority: it cannot forgive sins, pronounce blessing, give prophecy, or offer professional counsel of any kind. It is also instructed to decline crisis situations and point to real help — see Safety.

Our commitments

We label the experience as AI everywhere; we do not retain conversation content in our own database; the voice agent is configured not to save call audio and to delete conversation data on a zero-day schedule; we never let advertising interrupt a conversation in progress; and we ask you, sincerely, to bring whatever this tool stirs up to real people and real community. If any of these commitments ever appears broken, tell us via the contact page.