Privacy Policy

JCIL.AI is the operator of this platform, based at 130 Bishop Allen Drive, 5th Floor, Cambridge, MA 02139. JCIL stands for "Jesus Christ is Lord." It is a consumer product: a Scripture-aligned AI assistant built for individuals and families who want AI that shares their values. We do not sell institutional or enterprise accounts.

Your account, conversations and memory are stored on US infrastructure via the subprocessors listed in our Trust Center, and we do not move your stored data outside the United States. Answers from our primary model are generated through Anthropic's API, under Anthropic's commercial terms.

Account information. When you create an account we collect your email address, display name, hashed password (or a Google OAuth identifier), and account preferences. Google sign-in returns your name and email; no additional Google data.

Conversations. Messages you send and AI responses are stored so your conversation history is available when you return. Files you upload (images, PDFs, text files) are stored in encrypted storage and automatically deleted after 14 days.

Long-term memory. We extract a small number of preferences, recurring topics, and explicit feedback from your conversations to make future answers more personal. You can review and clear this memory at any time from your account settings.

Billing. For credit purchases we collect name, email, and payment method via Stripe. Card numbers never touch JCIL.AI servers. Stripe tokenizes everything.

Usage logs. Per-request metadata (which tool ran, token counts, cost in cents) so you can see your spend and we can catch abuse. Not shared with third parties beyond what's required to deliver the service.

Safety events. When the moderation pipeline detects a high-signal crisis pattern (suicide ideation, self-harm, abuse disclosure, child-at-risk), we record the event so it can be reviewed and appropriate follow-up can occur.

Automatic information. IP address, user agent, request timing, and similar request metadata. Used for rate limiting, abuse detection, and debugging. IP addresses and user agents are purged from our security audit records after 30 days.

• Deliver and improve the service you signed up for.
• Authenticate you, keep your account secure, and prevent abuse.
• Process payments and apply credits to your balance.
• Personalize future conversations via the long-term memory system you can control.
• Detect and respond to crisis situations flagged by the moderation pipeline, and screen images you attach to a chat for sexual, self-harm, or graphic-violence content before the assistant sees them (a refused image is deleted, and text messages are never sent to that screen).
• Meet legal obligations and respond to valid legal process.

We do not sell your data, rent it, share it for cross-context advertising, or use your conversations to train AI models. Anthropic, whose Claude models generate your answers and run the built-in web search, commits in its commercial terms not to train on your conversations; it retains API inputs and outputs for a limited period for abuse monitoring, and its privacy documentation has the specifics. The provider that hosts our backup model runs it in the United States with zero data retention on the endpoint we use. See the subprocessors page for the full list of services and what each one receives.

• Active conversations: kept as long as your account is active.
• File uploads: 14 days, then hard-deleted.
• Usage logs: 90 days, then deleted.
• Safety events: 180 days, then hard-deleted.
• Moderation logs: 30 days.
• Security violations: 24 months, then purged.
• Support tickets: 24 months, then purged.
• Shared chat links: expire and are deleted 90 days after they are created.
• Inactive accounts: we email a warning at 12 months without a sign-in, and delete the account and its data at 18 months.
• Billing transactions: 24 months (tax / audit requirement), then purged.
• Deleted accounts: 30-day grace window (recoverable), then permanent.
• Long-term memory: default 730 days (2 years). You can set it to 30, 90, 180, 365, or 730 days from account settings. There is no "keep forever" option.

Regardless of where you live, every US JCIL.AI user has the following rights. California residents are additionally protected by CCPA/CPRA.

• Right to know. Download a full export of your data (profile, conversations, memory, usage, billing) in JSON from account settings.
• Right to delete. Delete your account and associated data. Two options: scheduled (30-day window, reversible) or immediate (permanent hard-delete, irreversible).
• Right to correct. Update or correct any profile field from settings.
• Right to opt out of "sale" or "sharing." We don't sell or share personal information for cross-context behavioral advertising, so there is nothing to opt out of. A formal opt-out toggle is still available on the privacy settings page for CCPA compliance.
• Right to non-discrimination. Exercising your rights doesn't change the service you receive.

Do Not Sell or Share My Personal Information. We do not sell personal information, and we do not share it for cross-context behavioral advertising. To record a formal opt-out anyway, use the toggle on your privacy settings page or submit the compliance contact form with topic privacy / data rights.

To exercise any right, use the account settings page or submit the compliance contact form with topic privacy / data rights. We acknowledge within two business days and complete the request within 45 days.

JCIL.AI accounts are for people 16 and older. Someone younger may use JCIL only on a Family plan seat, invited by a parent or legal guardian who consents and manages the seat. JCIL.AI is not directed at children under 13. We do not knowingly collect personal information from anyone under 16 outside a managed Family seat, and we delete such accounts when we learn of them. Family managers can see time and activity for each seat on their plan, never conversations. If you believe a child has created an account outside these rules, contact us through the compliance form and we'll delete it.

Concrete measures, including TLS 1.3, AES-256 at rest, row-level security on every sensitive table, layered moderation, and vulnerability disclosure, are documented in full on our Security page.

We use a handful of first-party cookies for authentication and basic preferences. No cross-site tracking, no advertising pixels, no third-party analytics. We count our own funnel steps (a page viewed, an account created, a first message sent, a purchase made) in aggregate on our own servers, never tied to conversation content. Full details on our cookies page.

JCIL runs on Claude by Anthropic, an American AI company. We chose it deliberately, we align it with JCIL's convictions through our own instructions, and we test it continuously. Your conversations are never used for training, and your account, chats and memory are stored in the United States. Answers come from Claude models through Anthropic's API: Claude Haiku 4.5 by default, with Claude Sonnet 5 or Claude Opus 5 chosen per question when it deserves more. JCIL builds the product around it: the convictions, continuous quality testing (autonomous checks and human review), the memory, and the privacy design. If Claude is ever unavailable, JCIL switches to its backup, a frontier open-weight model hosted and processed in the United States, and tells you on screen. The backup answers only during an outage. Live web search runs through Anthropic's built-in search tool, and through a US search provider while the backup answers. Supporting service categories for infrastructure, billing, and delivery are listed in our subprocessor list. Asked whether you are talking to an AI, a bot, or a human, the assistant answers truthfully that it is an AI. It never claims to be human.

AI outputs are generated; they are not professional legal, medical, financial, or pastoral advice. Always verify material decisions with a qualified human.

We'll post the updated policy here and change the "Effective" date at the top. For material changes that reduce your rights, we'll notify active account holders by email at least 30 days before the change takes effect.

Privacy questions, data-rights requests, and legal process: use the compliance contact form.

This policy is one half of the agreement. The other half, covering accounts, billing, renewal, cancellation and acceptable use, is in our Terms of Service.

JCIL.AI
130 Bishop Allen Drive, 5th Floor
Cambridge, MA 02139
United States