AI Disclaimer

Trust scores, fraud flags, receipt extraction, and concierge answers are advisory.

• They are risk indicators, not proof of legitimacy or fraud.
• They are not legal, tax, medical, or investment advice.
• Final publish, payout, and compliance decisions are made by humans and automated policy rules — not by the language model alone.
• Donation amounts, allocations, and ledger hashes are computed by deterministic server code, not by AI.
• Models can err, be incomplete, or be targeted by prompt injection; grounding and tool checks reduce but do not eliminate risk.

Donors should apply independent judgment. Creators must still comply with all applicable laws. See also product limits in the repository documentation.