Review signals, human decisions, clear limits.
Fraud prevention is layered. No single social account, document, score, or AI model proves that a person or story is genuine.
Human review is required
AI can organize evidence and flag inconsistencies, but it cannot approve or reject a recipient by itself.
Private evidence stays private
Documents, payout information, risk reviews, and sensitive profile details are protected by database policies and staff-only access.
Scores are explained
Confidence indicators summarize reviewed signals such as identity, phone, payout, social links, and published updates. They are not guarantees.
Reports and appeals
Donors can report profiles or updates. Recipient decisions are logged and may be appealed through a documented moderation workflow.
Ready to see Blu in action?
Meet reviewed recipients and explore honest local impact estimates.