Synthetic prototypeNo real people, businesses, allegations or customer identifiers.

Human moderation

Publish the record.
Never manufacture a verdict.

Moderators decide eligibility, redaction and procedural fairness—not whether a person is morally trustworthy.

We can

Moderation actions

Request specific facts, context or better evidence
Redact private and unrelated third-party information
Reject ineligible subjects, threats, doxxing or manipulated evidence
Publish attributed claims after notice and response conditions
Correct, restrict, archive or remove under documented reason codes
Escalate high-risk cases for independent second approval

We cannot

Bright-line limits

Declare someone a fraudster based on a user report
Let AI publish, reject an appeal or impose a sanction
Hide a valid case because the subject pays
Publish raw IDs, addresses, phones or evidence files
Merge similar names or faces without verified identifiers
Allow the same moderator to provide both high-risk approvals