Trust & Compliance

JCIL.AI is built for Christians and their families, an audience where trust is the product. This is how we protect your data, manage our subprocessors, and stay honest about how the product works.

Our model, honestly

JCIL is powered by Claude from Anthropic, an American AI company built around safety research: everyday conversations run Claude Haiku 4.5, and demanding work may use larger Claude models. We did not build Claude and we do not pretend to: we chose it, and we name it here because you should know exactly whose model your family is talking to. What is ours is everything around it: the convictions, tested continuously against our quality bar, with autonomous checks and human review, for doctrine, pastoral safety and civic posture, the memory, and the privacy design. Every answer and every background helper that reads your words, the summaries, the suggestions, the routing, runs Claude on Anthropic's API; specialized heavy work, like building polished downloadable documents, may run a bigger Claude model (always Claude, never another company's). Your conversations are never used to train models, by us or by Anthropic, and your account, chats and memory are stored in the United States. If Claude has an outage, we say so plainly and the turn is free; your conversation is never quietly handed to some other company's model. The full list of every service that touches your data is on the subprocessors page.

Have a privacy or security question?

Ask us anything about your data, our security, or how the product works. One form, answered by a real person. Routes direct to our founder.

Contact us
Last updated: September 1, 2026.