Claims Compliance Checker

Overview

This skill screens consumer-facing product claims for regulatory compliance across the major U.S. and EU frameworks governing CPG products. It identifies prohibited claims, flags unsubstantiated language, and suggests compliant alternatives — preventing costly warning letters, listing removals, and brand reputation damage.

The checker operates as a pre-publication gate in the content workflow, sitting between copywriting and go-live approval.

When to Use

• Before publishing any PDP copy, packaging text, advertising creative, or social media content.
• When launching products in regulated categories (supplements, OTC, cosmetics, food, cleaning).
• After reformulating a product or changing ingredient sourcing.
• When expanding to new markets with different regulatory regimes (US → EU, US → Canada).
• During periodic compliance audits of existing content libraries.

Required Inputs

| Input | Description | Example | |---|---|---| | content | The full text to be validated | PDP copy, ad script, packaging text | | product_category | FDA/FTC product classification | "dietary supplement", "cosmetic", "conventional food", "OTC drug" | | approved_claims | List of claims cleared by legal/regulatory | ["Clinically tested", "Dermatologist recommended"] | | certifications | Active third-party certifications | ["USDA Organic", "Non-GMO Project Verified", "NSF Certified"] | | ingredient_list | Full ingredient/supplement facts panel | Structured list | | target_markets | Regulatory jurisdictions | ["US", "EU", "Canada"] | | channel | Distribution channel with its own policies | "Amazon", "Walmart", "Meta Ads", "Google Ads" | | substantiation_docs | Available clinical studies, tests, or certifications | List of document references |

Methodology

Step 1 — Claim Extraction & Classification

Parse the content and identify every explicit or implicit claim. Classify each into:

| Claim Type | Definition | Example | Governing Authority | |---|---|---|---| | Structure/Function | Describes role of nutrient in body | "Supports immune health" | FDA (DSHEA) | | Health Claim | Links substance to disease risk reduction | "May reduce risk of heart disease" | FDA (authorized/qualified) | | Drug Claim | Claims to diagnose, treat, cure, or prevent disease | "Cures acne" | FDA (prohibited for non-drugs) | | Nutrient Content | Quantifies nutrient level | "Excellent source of Vitamin C" | FDA (21 CFR 101.54) | | Cosmetic Claim | Affects appearance, not structure | "Reduces appearance of wrinkles" | FTC / FDA | | Environmental | Sustainability or eco-related | "Biodegradable", "Carbon neutral" | FTC Green Guides | | Comparative | References competitors or alternatives | "50% more effective than leading brand" | FTC (substantiation required) | | Testimonial | Consumer or expert endorsement | "9 out of 10 dentists recommend" | FTC Endorsement Guides |

Step 2 — Regulatory Rule Application

For each claim, apply the relevant regulatory framework:

FDA Framework (Food, Supplements, Cosmetics):

• Structure/function claims require the DSHEA disclaimer: "This statement has not been evaluated by the FDA. This product is not intended to diagnose, treat, cure, or prevent any disease."
• Health claims must be authorized (FDA-approved) or qualified (with required disclaimer language).
• Drug claims on non-drug products are automatic violations — flag as critical.
• Nutrient content claims must meet defined thresholds (e.g., "good source" = 10-19% DV per RACC).
• "Natural" on food: no FDA definition; flag for risk assessment. On cosmetics: must be substantiated.

FTC Framework (Advertising):

• All claims must be truthful, not misleading, and substantiated before dissemination.
• Comparative claims require head-to-head testing or equivalent evidence.
• Testimonials must reflect typical results or include clear disclaimers.
• "Clinically proven" requires at least one well-designed, peer-reviewed clinical trial.
• Influencer/endorsement content must disclose material connections (#ad, #sponsored).

EPA Framework (Household/Cleaning):

• Antimicrobial claims on surfaces require EPA registration.
• "Kills 99.9% of germs" requires specific EPA-registered test protocol reference.
• "Safer Choice" label requires active EPA partnership agreement.

EU Framework (if applicable):

• Health claims must appear on the EU Register of authorized claims (EC 432/2012).
• "Free from" claims governed by allergen labeling regulation (EU 1169/2011).
• Cosmetic claims regulated by EC 655/2013 common criteria.

Step 3 — Severity Scoring

Rate each flagged claim:

| Severity | Definition | Action | |---|---|---| | Critical | Illegal claim (drug claim on supplement, unauthorized health claim) | Block publication. Mandatory legal review. | | High | Unsubstantiated claim that could trigger FTC/FDA action | Rewrite required before publication. | | Medium | Technically compliant but misleading or missing required context | Add disclaimer or qualifying language. | | Low | Style preference or best-practice recommendation | Suggest improvement; not a blocker. |

Step 4 — Compliant Alternative Generation

For every flagged claim at medium severity or above, propose a compliant rewrite:

1. Downgrade drug claims to structure/function: "Cures joint pain" → "Supports joint comfort."
2. Add required qualifiers: "Clinically proven" → "Clinically tested in a 12-week study of 200 participants."
3. Replace absolute claims with hedged language: "Eliminates allergies" → "May help reduce common allergy symptoms."
4. Ensure environmental claims meet FTC Green Guides specificity: "Eco-friendly" → "Made with 80% post-consumer recycled plastic."

Step 5 — Channel-Specific Policy Overlay

Apply retailer/platform-specific restrictions on top of regulatory rules:

| Channel | Key Restrictions | |---|---| | Amazon | No disease claims in any field; no references to FDA approval; restricted categories require pre-approval | | Walmart | Prop 65 warnings required for CA; category-specific claim restrictions | | Meta/Google Ads | No before/after for supplements; restricted health categories require certification | | TikTok Shop | No medical claims; supplement ads require disclaimers in creative |

Output Specification

output:
  compliance_score: float           # 0-100 overall compliance rating
  total_claims_found: int
  claims_analysis: list[ClaimFinding]
  critical_violations: int
  high_violations: int
  medium_findings: int
  low_findings: int
  required_disclaimers: list[string]
  compliant_rewrites: dict          # Original → compliant alternative
  publication_recommendation: string # "approved" | "approved_with_changes" | "blocked"
  regulatory_references: list[string]

Analysis Framework

The Compliance Risk Matrix cross-references two axes:

• Likelihood of Enforcement (low/medium/high): Based on category scrutiny, recent enforcement trends, and claim visibility.
• Severity of Consequence (low/medium/high): Warning letter, product seizure, class-action exposure, retailer delisting.

Claims in the high-likelihood × high-severity quadrant are critical blockers. Claims in low × low are informational only.

Examples

Input: "Our probiotic gummy boosts immunity and prevents colds. Made with all-natural ingredients. #1 Doctor Recommended."

Findings:

1. "prevents colds" — Critical. Drug claim on a dietary supplement. FDA violation. → "Supports immune health†" + DSHEA disclaimer.
2. "boosts immunity" — Medium. Structure/function claim requires disclaimer notation (†). → Add DSHEA disclaimer.
3. "all-natural" — Medium. No FDA standard definition; FTC requires substantiation. → "Made with naturally sourced ingredients."
4. "#1 Doctor Recommended" — High. Requires substantiation (survey methodology, sample size). → "Recommended by doctors" with footnote to survey.

Compliance Score: 28/100. Recommendation: Blocked.

Guidelines

• Always err on the side of caution — false negatives (missed violations) are far more costly than false positives.
• This skill does not replace legal counsel. Critical and high findings must be routed to legal/regulatory teams.
• Maintain a living library of approved claim language per product category.
• Re-validate all content when regulations change (e.g., FDA final rule on "healthy," FTC updated endorsement guides).
• Track FDA warning letters and FTC consent decrees in your category as precedent signals.

Validation Checklist

• [ ] Every claim in the content is extracted and classified by type.
• [ ] Applicable regulatory frameworks (FDA, FTC, EPA, EU) are applied per product category.
• [ ] All critical and high violations have compliant alternative rewrites.
• [ ] Required disclaimers are identified and placed correctly.
• [ ] Channel-specific policy overlays are applied.
• [ ] Publication recommendation is clearly stated (approved / approved with changes / blocked).
• [ ] Regulatory references and citation links are provided for each finding.
• [ ] Substantiation documentation is cross-referenced against claim requirements.
• [ ] No drug claims exist on non-drug products.
• [ ] Environmental claims comply with FTC Green Guides specificity requirements.