Automated Copywriting Factuality and Claim Guardrail Matrix

Role: Head of Content Compliance and Quality Engineering specializing in high-velocity generative marketing pipelines.

Context

• Brand: {{brand_name}}
• Target Market: {{target_audience}}
• Marketing Channel & Format: {{campaign_type}}
• Regulatory Frameworks: {{regulatory_standards}}
• Blacklisted Claims: {{unsubstantiated_claims_list}}
• Verified Evidence Base: {{source_of_truth_repository}}

Task

Synthesize a production-ready factuality and claim validation guardrail framework that prevents {{brand_name}} from publishing hallucinated, unsubstantiated, or non-compliant marketing copy across {{campaign_type}}.

Method

1. Cross-reference {{regulatory_standards}} to establish mandatory disclosure requirements and banned superlatives.
2. Ingest {{unsubstantiated_claims_list}} and map corresponding semantic variants to block across generated copy.
3. Establish evidence-binding protocols that require every factual assertion to cite {{source_of_truth_repository}}.
4. Design a pre-generation prompt constraint schema that restricts generative temperature and speculative framing.
5. Create an automated post-generation semantic claim-verification checklist for {{target_audience}} copy.
6. Detail an automated red-flag scoring index that grades text drafts from zero to one hundred.
7. Outline remediation instructions for drafts scoring below the minimum automated release threshold.

Constraints

• MUST NOT permit absolute claims (e.g., 'always', 'guaranteed', 'the best') without verified evidence citation.
• Generated rules MUST enforce explicit disclosure placements based on {{campaign_type}} format limits.
• Guardrail validation stages MUST distinguish between stylistic edits and compliance blocks.
• All compliance rules MUST apply equally to headlines, body copy, and metadata.

Output format

Deliver the compliance system in three ordered modules:

1. Claim Verification Logic (flowchart-style numbered rules specifying check order and pass/fail logic)
2. Regulatory Boundary Table (columns: Claim Type, Disallowed Phrases, Compliant Alternative Pattern, Mandatory Disclaimer)
3. Automated Quality Gate Scoring Rubric (scoring brackets from 0 to 100 with clear release criteria, max 300 words)

Self-review

• Ensure all rules directly reflect the boundaries imposed by {{regulatory_standards}}.
• Validate that the scoring rubric provides unambiguous pass/fail criteria.
• Check that every blacklisted concept from {{unsubstantiated_claims_list}} has a concrete compliant alternative.