Attribution & AI Visibility Disclosure Policy

1. Purpose

This policy ensures that authorship, AI involvement, and synthetic content are always visible and traceable. Attribution and AI visibility are core to maintaining trust and accountability in digital ecosystems.

By making origin and AI involvement transparent, we enable informed decisions, support legal claims, prevent fraud, and maintain the integrity of creative and scientific work.

2. Attribution Requirements

A. Authorship Declaration

Every artifact registered on the platform must include:

• Primary author(s) or creator(s)
• Creation date and time
• Organization or institutional affiliation (if applicable)
• License or usage rights

B. Attribution Immutability

Attribution information becomes part of the immutable provenance record. Once recorded, it cannot be deleted or retroactively modified. Corrections create new lineage entries rather than altering original records.

C. Derivative Works

Derivative works must maintain attribution to original sources, including:

• Links to source artifacts
• Original author credits
• License continuity
• Lineage references

3. AI Involvement Disclosure

A. Types of AI Involvement

Users must disclose when content is:

• AI-Generated: Wholly created by AI systems without substantial human input
• AI-Assisted: Created with human input but using AI tools (e.g., AI writing assistants, code generators)
• AI-Enhanced: Human-created content improved or modified by AI (e.g., image enhancement, translation)
• AI-Trained: Content used to train AI models (with user consent)

B. Required Disclosure Information

When AI is involved, users must provide:

• Type of AI involvement
• Model name and version (if known)
• Training data sources (if applicable)
• Human oversight level
• Extent of AI contribution

C. Automated Detection

The platform may automatically detect synthetic content using fingerprinting, statistical analysis, and pattern recognition. Detected AI involvement is flagged in provenance records even if not explicitly declared.

Note: Detection is not perfect. Users are still responsible for accurate disclosure.

4. Synthetic Content Visibility

A. Visibility Signals

Synthetic content is marked with visibility signals that include:

• Visual indicators in user interfaces
• Metadata tags in provenance records
• API flags for programmatic access
• Export metadata in downloadable formats

B. Downstream Visibility

AI involvement and synthetic signals are preserved throughout the provenance chain. Anyone viewing or using the content can see:

• Original AI involvement declarations
• Detection results (if applicable)
• Model lineage and training data
• Human oversight information

C. Immutable Visibility

AI visibility metadata becomes part of immutable provenance records. It cannot be removed, hidden, or retroactively modified. This ensures permanent transparency.

5. Model & Dataset Attribution

A. Model Attribution

When AI models are used, provenance records include:

• Model identifier and version
• Model passport information
• Training dataset references
• License and attribution requirements

B. Dataset Attribution

When datasets are used for training, provenance records include:

• Dataset identifier and version
• Dataset passport information
• Source data provenance
• License and usage rights

6. Prohibited Conduct

Users may not:

• Misrepresent authorship (claiming others' work as your own)
• Hide or falsify AI involvement
• Remove or obscure attribution metadata
• Present AI-generated content as human-created
• Break lineage or derivative attribution chains
• Use the platform to "launder" AI content as original work

Violations may result in content removal, account suspension, or legal action.

7. Enforcement Mechanisms

A. Automated Detection

The platform automatically:

• Detects synthetic content patterns
• Flags missing or inconsistent attribution
• Validates lineage continuity
• Monitors for tampering with metadata

B. Human Review

Reports of misattribution or hidden AI involvement trigger human review, which may include:

• Forensic analysis of provenance records
• Verification of authorship claims
• Investigation of AI detection signals
• Correction or enforcement actions

C. Corrective Actions

When violations are found, we may:

• Add correction entries to provenance records
• Flag content with warnings
• Remove non-immutable content
• Suspend or terminate accounts
• Pursue legal action in cases of fraud

8. Downstream Use & Attribution

A. Export Formats

When content is exported, attribution and AI visibility information is included in:

• Provenance manifests
• Metadata files
• Watermarks or embedded signals (where applicable)
• API responses

B. Third-Party Use

Third parties using content from the platform must respect attribution and license requirements as recorded in provenance records. The platform provides tools to verify and enforce these requirements.

9. Limitations & Disclaimers

While we strive for accuracy, the platform cannot guarantee:

• Perfect AI detection (some synthetic content may go undetected)
• Verification of authorship claims (we rely on user declarations)
• Enforcement outside the platform (attribution requirements apply within our system)
• Legal outcomes (provenance records support but do not guarantee legal claims)

10. Updates

This policy may be updated to reflect changes in AI detection capabilities, attribution standards, or platform features. The "Last Updated" date reflects the most recent revision.