PRIV ProtocolPRIV Docs
Data Bounty Board

Content Rules and Guidelines

Mandatory rules for all Data Bounty Board submissions including ownership, quality standards, and prohibited content.

Overview

All submissions to the Data Bounty Board must comply with these mandatory rules. Violations result in rejection, reputation penalties, and potential account suspension.

Repeated violations of content rules will result in permanent ban from the Bounty Board and forfeiture of pending earnings.

The 10 Mandatory Rules

Rule 1: Original Content Only

You must own or have full rights to all submitted content.

AllowedNOT Allowed
Photos you personally tookStock photos
Videos you recordedClips from movies/TV
Voice recordings of yourselfRecordings of others without consent
Text you wroteCopied articles or AI-generated text
Audio you created/recordedMusic with copyright

Evidence requirements:

  • Be prepared to provide proof of creation (RAW files, timestamps, etc.)
  • If using licensed content, provide license documentation
  • If including third-party elements, provide release forms

Rule 2: No AI-Generated Content

Content generated by AI tools is strictly prohibited.

This includes:

  • Images from Midjourney, DALL-E, Stable Diffusion, etc.
  • Text from ChatGPT, Claude, or other LLMs
  • Audio/voice from ElevenLabs, Synthesia, etc.
  • AI-enhanced or AI-upscaled content
  • AI "style transfer" modifications

Why this matters:

  • AI-generated content creates legal uncertainties
  • Bounty creators need authentic human-created data for training
  • AI detection tools are used to screen submissions

We use multiple AI detection methods. Attempting to disguise AI content as human-created will result in immediate permanent ban.

Rule 3: No Personally Identifiable Information (PII)

Never include PII of yourself or others.

Prohibited PII includes:

CategoryExamples
IdentityFull names, ID numbers, SSN, passport numbers
ContactPhone numbers, email addresses, home addresses
FinancialBank accounts, credit cards, tax IDs
BiometricFingerprints, retina scans, DNA data
HealthMedical records, prescriptions, diagnoses
CredentialsPasswords, PINs, security questions

Handling faces in images/videos:

  • Blur or obscure all recognizable faces
  • Use "incidental" exception only for distant crowd shots
  • Get written consent if face must be visible

Rule 4: No Illegal Content

Content depicting or promoting illegal activities is prohibited.

This includes:

  • Drug manufacturing or distribution
  • Weapons (especially illegal modifications)
  • Fraud schemes or financial crimes
  • Hacking tutorials or malware
  • Child exploitation (ZERO TOLERANCE)
  • Terrorism or extremist content
  • Animal cruelty
  • Harassment or stalking

Child Sexual Abuse Material (CSAM) is reported to law enforcement immediately. There is no appeal process.

Rule 5: No Explicit Adult Content

Sexual or explicit content is not permitted.

Prohibited content:

  • Nudity (except artistic/medical contexts where specifically requested)
  • Sexual acts or simulation
  • Pornographic material
  • Sexual solicitation
  • Sexualized content involving minors (ZERO TOLERANCE)

Exception: Some medical/anatomy bounties may allow clinical nudity with explicit consent and verification.

Rule 6: No Hate Speech or Discrimination

Content promoting hate based on protected characteristics is banned.

Protected characteristics include:

  • Race and ethnicity
  • Religion
  • Gender and gender identity
  • Sexual orientation
  • Disability
  • National origin
  • Age

This applies to:

  • Slurs and derogatory terms
  • Symbols of hate groups
  • Discriminatory statements
  • Dehumanizing content

Rule 7: No Violence or Gore

Graphic violent content is prohibited unless specifically requested.

Prohibited:

  • Real violence against humans or animals
  • Graphic injury or death
  • Torture or abuse
  • Self-harm promotion

Exception: News/documentary bounties may accept sanitized violence with explicit labeling.

Rule 8: No Misinformation

Factually incorrect content intended to deceive is prohibited.

This includes:

  • Fake news articles
  • Manipulated media (deepfakes)
  • Health misinformation
  • Election misinformation
  • Conspiracy theories presented as fact

For creative/fictional content:

  • Clearly label as fiction
  • Don't mimic news formats
  • Don't impersonate real people

Rule 9: Quality Standards Must Be Met

All content must meet minimum quality thresholds.

Images

CriterionRequirement
ResolutionMinimum 100x100 pixels
FocusMain subject must be sharp
ExposureProperly lit, not blown out or underexposed
ArtifactsNo heavy compression artifacts
WatermarksNo watermarks or overlays

Videos

CriterionRequirement
ResolutionMinimum 240p
StabilityNo excessive shake (unless intended)
Audio syncIf audio present, must be synced
CompletenessNo cut-off or corrupted segments

Audio/Voice

CriterionRequirement
ClaritySpeech must be understandable
NoiseMinimal background noise
DurationWithin specified limits
FormatCorrect sample rate

Text

CriterionRequirement
GrammarReasonably correct for the language
CoherenceLogically structured
RelevanceOn-topic for the bounty
FormattingFollows specified format (JSON, etc.)

Rule 10: Accurate Metadata

All metadata must be truthful and accurate.

Required accuracy:

  • File attributes (resolution, duration, etc.)
  • Location data (if provided)
  • Timestamps and dates
  • Language identification
  • Content descriptions

Prohibited:

  • Fabricated metadata
  • Spoofed EXIF data
  • Misleading descriptions
  • Category misclassification

What IS Allowed

To clarify, here is content that IS acceptable:

CategoryExamples
Original photographyYour own photos of any legal subject
Self-recordingsVoice samples, videos of yourself
Creative writingStories, conversations, articles you wrote
Licensed contentContent you have commercial rights to
Public domainWorks with expired copyright (verify first)
Anonymized dataProperly anonymized behavioral/location data
Non-offensive contentSafe for work content in all categories

Privacy Requirements

For Images and Videos

Faces: Must be blurred, pixelated, or obscured unless:

  • The person is you AND you consent
  • You have written release from the person
  • The bounty explicitly allows faces AND you have consent

License plates: Must be blurred or obscured

Street addresses: Must be blurred if clearly visible

Private property: Avoid if owner would object

For Text and Behavioral Data

Remove or anonymize:

  • Names (use placeholders like [NAME])
  • Email addresses
  • Phone numbers
  • User IDs that can be traced
  • IP addresses
  • Specific locations more precise than city

For Voice Recordings

Your voice only unless you have:

  • Written consent from all speakers
  • Proof of consent on file
  • Bounty specifically allows multiple speakers

Quality Standards by Type

Image Quality Tiers

TierResolutionFile SizeUse Case
Basic100-799pxUp to 1MBThumbnails, previews
Standard800-1919pxUp to 5MBGeneral training
HD1920-3839pxUp to 15MBHigh-quality training
Premium3840px+Up to 25MBProfessional applications

Video Quality Tiers

TierResolutionDurationUse Case
Basic240-479p1-60 secQuick clips
Standard480-719p1-300 secGeneral training
HD720-1079p1-600 secQuality training
Full HD1080-2159p1-1800 secProfessional
4K2160p+1-3600 secPremium applications

Audio Quality Tiers

TierSample RateDurationUse Case
Basic8-15 kHz1-300 secSimple speech
Standard16-44 kHz1-1800 secGeneral audio
HD44-96 kHz1-3600 secHigh quality
Studio96-192 kHz1-7200 secProfessional

Content Moderation Process

Automated Screening

All submissions pass through automated checks:

Submission


┌─────────────────┐
│ Format Check    │ ← Correct file type, size, resolution
└────────┬────────┘


┌─────────────────┐
│ Duplicate Check │ ← SHA-256 hash comparison
└────────┬────────┘


┌─────────────────┐
│ AI Detection    │ ← Checks for AI-generated content
└────────┬────────┘


┌─────────────────┐
│ NSFW Detection  │ ← Flags explicit content
└────────┬────────┘


┌─────────────────┐
│ Quality Score   │ ← Assesses image/audio quality
└────────┬────────┘


    Approved / Flagged / Rejected

Approver Review

Flagged content goes to staked approvers who:

  • Review against bounty requirements
  • Check for rule violations
  • Assess quality subjectively
  • Cast votes toward approval or rejection (3-vote quorum, 66.67% consensus)

Violation Consequences

SeverityViolationsConsequence
WarningMinor quality issuesRejection, no penalty
MinorRepeated quality issues-5 reputation, 1-day cooldown
ModeratePII violations, misrepresentation-20 reputation, 7-day cooldown
SevereHate speech, explicit content-50 reputation, 30-day suspension
Permanent BanCSAM, fraud, repeated severeAccount termination, wallet blacklist

Appeal Process

If your submission is rejected and you believe it complies with rules:

  1. Review the rejection reason - Check the specific violation cited
  2. Gather evidence - Proof of ownership, original files, etc.
  3. Submit appeal - Via dashboard within 7 days
  4. Wait for review - Appeals reviewed within 48 hours

Appeals are NOT available for:

  • CSAM (reported to law enforcement)
  • Clear fraud attempts
  • Repeat offenders on final warning

FAQ

Can I submit content with my face visible?

Only if the bounty explicitly allows it AND you're the only person visible. Most bounties require face blurring.

What if I accidentally include PII?

If caught before approval, your submission is rejected with a warning. Learn to check more carefully. If discovered after approval, more serious consequences apply.

How do I know if content is AI-generated?

If you used any AI tool to create, enhance, or modify the content, it's considered AI-generated. This includes upscaling, style transfer, and "enhancement" tools.

Can I use Creative Commons content?

Only if the CC license allows commercial use (CC0, CC-BY, CC-BY-SA). You must provide license documentation.

What about screenshots?

Screenshots of your own content are allowed. Screenshots of others' content, apps, or websites require specific permission.

Next Steps