Security & Governance Standard

The Trust & Escrow Architecture

How EarnSphere eliminates the six structural trust violations of the rewards industry through cryptographic verification, escrow locks, and auditable accounting.

Smart Escrow Budgeting

Requesters must pre-fund 100% of the reward budget into platform escrow before any task or campaign goes live. Once a worker accepts a task slot, the reward is locked in escrow and cannot be withdrawn or canceled by the requester.

Strict 72-Hour Auto-Approval SLA

To prevent infinite pending holds, all submissions are bound by an automated 72-hour countdown. If a requester does not actively review a submission within 72 hours, the platform daemon automatically triggers approval and credits the worker’s balance.

Immutable Append-Only Ledger

Balances on EarnSphere are not raw mutable numbers that can be arbitrarily edited or lost during database updates. Every balance change is an immutable double-entry ledger event protected by database engine triggers that prohibit UPDATE and DELETE statements.

Mandatory Rejection Taxonomies & Appeals

Requesters cannot reject submissions with vague or missing reasons. They must select from our standardized taxonomy (e.g. Incomplete Proof, Invalid Screenshot, Unmet Criteria) and provide explanatory notes. Workers have a 24-hour window to contest rejections.

Worker Trust Score (300 to 850 Index)

Our FICO-style reputation score evaluates attention check pass rates, on-time submissions, and historical approvals. High-trust workers (750+) unlock instant zero-wait payouts and priority access to high-paying AI evaluation tasks.

Prorated Drop-off Protection

If a third-party partner unexpectedly terminates a survey session due to quota caps after minute 3, EarnSphere automatically compensates the participant with prorated drop-off points funded by our platform reserve.

Standardized Governance

Standardized Rejection Taxonomies

Requesters cannot arbitrarily reject work. All rejections must be accompanied by an audit-validated reason code:

REJ_PROOF_INCOMPLETESubmission lacks required proof fields, transaction IDs, or complete answer forms.
REJ_INVALID_SCREENSHOTUploaded screenshot is cropped, unreadable, or missing the verification timestamp.
REJ_CRITERIA_UNMETWorker failed to follow device criteria (e.g. tested on desktop instead of physical iPhone).
REJ_LOW_QUALITYWritten qualitative response failed minimum length, detail, or quality criteria.