# Usage: curl -sSL https://seed.show/marketing.ftc.compliance | bash -s <install-path>
# <install-path> is the directory where the file should land.

set -euo pipefail
[ -z "${1:-}" ] && {
  echo "install requires a path: curl -sSL https://seed.show/marketing.ftc.compliance | bash -s <install-path>" >&2
  exit 1
}
TARGET="$1"
mkdir -p "$TARGET"
DEST="$TARGET/seed-fold.6XDC6o.folded.md"

cat > "$DEST" <<'PORTDOWN_A0CBD61B'
<!--
This is a `.folded.md` archive — a directory packed into one markdown
file. The markers below are load-bearing; don't edit them directly.

To unpack (requires bash — if you have no shell, tell the user):
  1. curl -sSLf https://fold.dom.vin/skill | bash -s <INSTALL_DIR>
  2. <INSTALL_DIR>/fold/scripts/unfold <this-file>
     (or: unfold <this-file>  if fold/scripts is on your PATH)
-->

---
fold: true
marker: b41618
at: 2026-05-07T16:16:17Z
root: seed-pack.q2qExB
---

<!--fold:b41618@file path="README.md" mode="644"-->
# FTC Marketing Compliance Context

## What this is and what it is not

This seed gives an AI agent working on marketing copy, campaign strategy, or advertising review a structural map of FTC compliance requirements. It covers the stable framework — Section 5 authority, the deceptive practices standard, key regulations, and disclosure mechanics.

**Do not use this as a substitute for legal review.** Advertising compliance questions with dollar exposure, enforcement risk, or novel fact patterns require a qualified advertising attorney. The FTC's enforcement posture, penalty amounts, and guidance documents are updated; this seed holds the framework, not the current enforcement temperature. Fetch sources.md for live document URLs.

---

## Mental model: FTC authority is two-pronged

The FTC's enforcement power over marketing rests on two distinct foundations:

**Prong 1 — Unfair acts or practices**: A practice is unfair if it (1) causes or is likely to cause substantial consumer injury, (2) that is not reasonably avoidable by consumers, and (3) is not outweighed by countervailing benefits. The FTC uses the unfairness prong for dark patterns, negative option billing traps, and discriminatory algorithmic outcomes.

**Prong 2 — Deceptive acts or practices**: A representation, omission, or practice is deceptive if it (1) is likely to mislead a reasonable consumer acting reasonably under the circumstances, and (2) is material — meaning it likely affects purchasing decisions. The FTC does not need to show actual consumer harm; likelihood is enough. Materiality is presumed for health, safety, and cost claims.

**Most marketing violations are deception cases.** The question is almost always: does this representation, taken as a whole and in context, create a false impression in a reasonable consumer's mind, about something that matters to their decision?

The FTC also administers specific regulations that carry civil penalties independent of Section 5 enforcement. Key ones: Endorsement and Testimonial Guides (16 CFR Part 255), CAN-SPAM Act rules, COPPA (16 CFR Part 312), and the Negative Option Rule (16 CFR Part 425).

---

## What agents get wrong

### 1. Influencer disclosure requirements

The FTC's Endorsement Guides (updated 2023, 16 CFR Part 255) require that **material connections** between an endorser and a brand be clearly and conspicuously disclosed. A material connection is anything that might affect the weight consumers give to the endorsement: payment, free products, discounts, affiliate commissions, employment relationship, family relationship, or any other financial or personal tie.

Errors agents make:
- Draft influencer briefs describing deliverables (posts, stories, links, swipe-ups) without specifying disclosure requirements for each placement type
- Assume "#ad" or "#sponsored" buried in a list of hashtags is sufficient — the 2023 guides explicitly require disclosures to be **hard to miss** and adjacent to the claim, not in a profile bio or a link-in-bio
- Treat organic-looking posts as compliant because no cash changed hands — free product, gifted items, and affiliate commissions all create material connections requiring disclosure
- Fail to flag that **employee posts about their employer** require disclosure of the employment relationship, even on personal accounts
- Assume the brand's compliance obligation ends at the brief — the FTC holds brands liable for influencer disclosures they control, enable, or fail to correct
- Miss that the 2023 revision explicitly added **virtual influencers** (AI-generated personas) and requires disclosure that the "person" is a computer-generated character

Three-part test for adequate disclosure: **Clear** (unambiguous language — "Ad:", "Paid partnership with [Brand]" — not "collab" or "partner"), **conspicuous** (visible without scrolling or expanding, not in fine print), **close to the claim** (adjacent to the endorsement, not separated by other content).

### 2. Claim substantiation

Any objective claim — "fastest," "clinically proven," "#1 dermatologist recommended," "up to 50% savings," "users lost an average of X pounds," "3x more effective" — requires a **reasonable basis** before the claim is published, not assembled after a complaint arrives.

Errors agents make:
- Draft ad copy with performance claims ("boosts conversion by 3x," "dermatologist tested," "reduces wrinkles in 2 weeks") without confirming what evidence exists
- Treat hedging language ("may help," "some users report," "results may vary") as a safe harbor — hedging reduces but does not eliminate the substantiation requirement for the underlying implied claim
- Confuse puffery with objective claims — puffery ("amazing," "you'll love it," "the best coffee you've ever had") is generally non-actionable because no reasonable consumer takes it as a specific verifiable assertion; objective comparative claims are not puffery
- Assume studies from other products transfer to this product — the substantiation must apply to the specific formulation, dosage, and population being marketed
- Miss the elevated standard for health claims: **competent and reliable scientific evidence** (generally randomized controlled trials for drug-like efficacy claims; lower bar for structure/function claims)

Rule: **have the evidence before you publish.** The FTC's standard is what substantiation a reasonable marketer would have had before making the claim.

### 3. Negative option and free trial traps

The FTC's Negative Option Rule (2023 update, 16 CFR Part 425) covers any offer where the seller interprets consumer silence or inaction as consent to be charged. Applies to subscriptions, auto-renewals, free-to-paid trial conversions, and continuity plans.

Requirements under the 2023 rule:
- **Disclose clearly and conspicuously** all material terms *before* obtaining billing information: price, billing frequency, cancellation terms, deadline to avoid the next charge. Not in fine print, not on a subsequent screen, not in the terms of service
- **Get affirmative consent** to the negative option feature — a separate checkbox or affirmative act, distinct from consenting to the service itself; pre-checked boxes do not qualify
- **Simple cancellation** — the cancellation mechanism must be at least as easy as sign-up; if sign-up is one click online, cancellation cannot require a phone call, an email, a chat, or a waiting period

Errors agents make:
- Design onboarding flows that pre-check the auto-renew option or bury renewal terms below the payment button
- Build "free trial" flows without displaying, adjacent to the CTA, when the paid subscription begins and the exact charge amount
- Accept cancellation flows that require contacting support when sign-up was self-serve
- Miss that the FTC's Operation Dark Patterns enforcement treats friction-adding cancellation flows (mandatory retention offers, multi-step "are you sure" flows, phone-only cancel) as Section 5 violations regardless of what the terms of service technically disclose

### 4. AI-generated content and fake reviews

The FTC's fake review prohibition (final rule, October 2024) explicitly covers AI-generated reviews and testimonials. Civil penalties apply per violation.

Prohibited under the 2024 Fake Reviews Rule:
- Buying, creating, or disseminating **fake consumer reviews** — including reviews generated by AI tools
- **Insider reviews** without clear disclosure — employees, company insiders, or company-controlled personas posting reviews without identifying their connection
- **Review suppression** — selectively publishing only positive reviews when you've solicited a broader set
- **Review hijacking** — repurposing reviews from a different product, formulation, or period

Operation AI Comply (September 2024) produced five enforcement actions against companies using AI to generate fake reviews, fake testimonials, and deceptive weight-loss claims. The underlying violations were Section 5 deception — the AI origin was aggravating, not the distinct charge.

Errors agents make:
- Generate marketing testimonials, "customer success stories," or review content using AI and present it as real consumer experience
- Apply the disclosure requirement only to political AI-generated content (state laws) when the FTC's fake review rule covers commercial contexts
- Assume that AI-generated *marketing copy* (which does not claim to be a real consumer) is the same compliance category as AI-generated *testimonials* (which do claim to be real consumer experience) — they aren't

### 5. CAN-SPAM vs. CASL — jurisdiction matters for email

**CAN-SPAM** (US federal, 2003): An **opt-out** regime. Commercial email is permitted without prior consent. Requirements: no deceptive subject lines or headers, physical postal address, identification as an advertisement (if not reasonably obvious), and a functioning opt-out mechanism honored within **10 business days**. Applies to any commercial email sent to a US recipient; the sending company is liable for third-party senders (ESPs, agencies) if they knew or consciously avoided knowing the third party would violate CAN-SPAM.

**CASL** (Canada, 2014): An **opt-in** regime. Commercial electronic messages require **express or implied consent** before sending. Express consent: recipient affirmatively opted in. Implied consent: existing business relationship — purchase within the prior 24 months, or inquiry/application within the prior 6 months. Applies to any commercial electronic message sent to or from Canada. Covers SMS, instant messaging, and other commercial electronic messages in addition to email. Civil penalties: up to CAD $1 million per violation for individuals, CAD $10 million per violation for organizations.

Errors agents make:
- Apply CAN-SPAM opt-out logic to Canadian recipients (CASL requires opt-in consent; no consent = no send)
- Treat a past Canadian purchase as ongoing implied consent after the 24-month window expires
- Miss that a Canadian *inquiry* (quote request, form fill, support contact) creates only 6-month implied consent, not 24-month
- Design SMS campaigns using email consent records — CASL covers SMS as a commercial electronic message; US TCPA adds additional consent requirements for US SMS

---

## What AI is changing

The FTC has named AI-enabled deception a top enforcement priority. Several distinct legal questions are in motion simultaneously:

**Fake reviews and testimonials**: The 2024 Fake Reviews Rule explicitly covers AI-generated fake reviews. This is settled enforcement territory: using AI to generate fake consumer reviews is a per-violation penalty-eligible violation. No ambiguity.

**Voice cloning in advertising**: The FTC has flagged AI voice cloning as an emerging deception issue — using a synthesized version of a real person's voice to imply endorsement without consent. Several states (California, New York, others) have right-of-publicity statutes that cover voice; the FTC's Section 5 analysis would focus on whether the cloned voice creates a false impression of endorsement. No specific FTC rule yet; active enforcement risk via Section 5 and state law.

**Discriminatory algorithmic advertising**: The FTC and DOJ have brought actions under fair housing and equal opportunity laws targeting ad-targeting systems that exclude protected classes. Meta's housing ad targeting was a direct enforcement precedent. Advertisers using lookalike audiences, exclusion lists, or behavioral targeting for credit, housing, employment, or insurance verticals face heightened scrutiny.

**Disclosure of AI in advertising**: No current FTC rule requires advertisers to affirmatively disclose that marketing copy, images, or video were created by AI *unless* that omission would be materially misleading — for example, using an AI-generated "before/after" image implies a real human result. California has a pending disclosure requirement for synthetic media in advertising. Several states require disclosure of AI-generated content in political advertising.

**FTC's "AI" enforcement lens**: The FTC has stated it will not create AI-specific exemptions — existing Section 5 deception analysis applies regardless of the content's origin. The consequence: if a human couldn't say it, an AI can't say it either. AI-generated content inherits all existing substantiation, disclosure, and non-deception obligations.

**Practical guidance for advertisers**:
- AI-generated testimonials that present as real consumer experience: prohibited under the fake reviews rule; no safe harbor for AI origin
- AI-generated images used in before/after comparisons: require the same substantiation as photography of real results
- AI chatbots used in sales flows: subject to the same deceptive practices standard as human representatives — deceptive statements by a chatbot are Section 5 violations
- AI tools used to generate marketing claims: the advertiser remains responsible for substantiation; the AI output does not create substantiation

---

## Key stable facts

**FTC Act Section 5 standard**: A practice is deceptive if it (1) is a representation, omission, or practice, (2) likely to mislead consumers acting reasonably under the circumstances, and (3) is material. "Likely to mislead" does not require proof of actual harm — the FTC needs only the tendency.

**Three-part disclosure test**: Applies across endorsements, testimonials, material connections, and any other required disclosure. Disclosures must be: **Clear** (plain language, not legalese or ambiguous shorthand), **Conspicuous** (easy to notice, read, and understand — not buried in fine print, hashtag stacks, or below the fold), and **Close to the claim** (adjacent to the specific representation being qualified, not in a footer, bio, or general disclaimer elsewhere).

**Civil penalty exposure**: Section 5 alone historically produced equitable relief, not per-violation penalties. However: (1) after a cease-and-desist order, subsequent violations carry penalties up to $51,744 per violation per day; (2) FTC rules promulgated under Section 18 carry direct civil penalties for first violations — the Negative Option Rule and Fake Reviews Rule are both Section 18 rules; (3) state "mini-FTC Acts" frequently allow private rights of action with statutory damages.

**COPPA age threshold**: The Children's Online Privacy Protection Act applies to online collection of personal information from children under 13. Requires verifiable parental consent before collection. Applies to operators directed to children or with actual knowledge of under-13 users. Penalties up to $51,744 per violation (2023 adjustment). The FTC is currently reviewing COPPA for a potential update to the age threshold and to address AI systems.

**State mini-FTC Acts**: Every US state has a consumer protection statute modeled on Section 5. Most prohibit unfair or deceptive acts or practices. Many allow private rights of action (unlike Section 5, which is FTC-enforcement-only). California's UCL, False Advertising Law, and CLRA are the most frequently litigated. Multi-state campaigns must account for the most restrictive applicable state law.

---

## Before any marketing review

Fetch sources.md for current FTC guidance URLs. The FTC updates its business guidance documents, enforcement priorities, and rulemaking; always pull the current version rather than relying on descriptions in this file. Check the FTC Cases and Proceedings database for enforcement actions in the relevant vertical from the past 18 months — enforcement posture shifts faster than the rules themselves.

See failure-modes.md for the most common FTC compliance failures, their enforcement pattern, and what compliant practice looks like.
<!--fold:b41618@file path="failure-modes.md" mode="644"-->
# FTC Marketing Compliance: Common Failure Modes

Seven failure patterns account for the majority of FTC marketing enforcement actions. Each entry covers the violation pattern, how the FTC typically discovers and pursues it, and what compliant practice looks like.

---

## 1. Undisclosed paid endorsements

**The violation pattern**: An influencer, affiliate, reviewer, or employee posts positive content about a brand without disclosing a material connection — payment, free product, discount, affiliate commission, employment, family relationship. The content reads as an organic opinion. The brand knew about the post (or directed it) and did not ensure a disclosure was made.

**What "material connection" actually covers**: Cash payment is the obvious case. The FTC's 2023 Endorsement Guides make clear that material connections also include: gifted products (regardless of value), loaned products, discounts on future purchases, affiliate commissions, employment relationship, professional service relationship (consulting, board seat), and close personal relationships (family members, close friends). "I just like the product and wanted to share" is not a defense if any of these relationships exist.

**Enforcement pattern**: The FTC typically issues warning letters to influencers and brands in bulk (the 2017 warning letters to 90+ influencers, the 2020 warning letters during COVID supplement marketing). Escalation produces consent decrees with monitoring, reporting requirements, and civil penalty exposure for future violations. Since the 2023 rules, the FTC has made clear it will pursue brands — not just endorsers — for inadequate disclosure programs.

**What compliant looks like**:
- Brand brief explicitly states disclosure requirements for every placement type (feed post, story, reel, video, blog, podcast mention) with approved language
- Disclosure language is specific and prominent: "Ad:", "Paid partnership with [Brand]", "#sponsored" — not "collab," "partner," "gifted" without "ad," or hashtag-buried labels
- Disclosure appears at the beginning of the caption (not after the "more" truncation), in the first seconds of video, or in the image itself — never only in a profile bio or link-in-bio
- Brand has a documented review process to verify disclosures before posts go live, and a correction process if disclosures are missing

---

## 2. Unsubstantiated advertising claims

**The violation pattern**: An ad, website, or marketing piece makes a specific, objective claim — a performance number, a health benefit, a comparative ranking, a scientific endorsement — without the advertiser having collected evidence that supports the claim before publishing it.

**What counts as an objective claim**: "Clinically proven to reduce wrinkles by 30% in 4 weeks." "Recommended by 9 out of 10 dentists." "Our supplement users lost an average of 14 pounds in 30 days." "#1 customer-rated pest control service." "EPA-registered." "Kills 99.9% of germs." Each of these asserts a specific, verifiable fact. Puffery ("the best coffee in the world," "you'll feel amazing") does not require substantiation because no reasonable consumer understands it as a specific factual claim; the line is crossed when the statement implies a verifiable assertion.

**Enforcement pattern**: Unsubstantiated health and weight-loss claims are a perennial FTC target (Operation Cure.All, 2000s; continuing through 2024 Operation AI Comply which included AI-generated weight-loss claims). Enforcement typically involves a consent order requiring the advertiser to have a specified evidentiary standard before making future claims (often "competent and reliable scientific evidence" for health claims, meaning randomized controlled trials). Civil penalties for subsequent violations.

**What compliant looks like**:
- Claims are reviewed against existing evidence before publication; a legal or regulatory review step is built into the approval workflow for objective performance claims
- For health efficacy claims: randomized controlled trials on the specific formulation and population being marketed; published studies on related compounds or competitor products do not transfer
- Survey-based claims ("9 out of 10 dentists recommend...") require a methodologically sound survey (representative sample, neutral question wording, full methodology disclosed on request); the asterisk must link to a description of the survey
- "Up to X%" savings claims require that the "up to" figure is genuinely achievable and that the baseline (the comparison price) is the real prevailing price, not an inflated reference price

---

## 3. Deceptive pricing

**The violation pattern**: A price is advertised in a way that creates a false impression of value, discount, or savings. The most common forms: (1) fictitious reference prices ("Was $200, now $79" when $200 was never a real selling price), (2) drip pricing (advertising a base price and adding fees at checkout — processing fees, service fees, resort fees — that were not disclosed upfront), (3) bait-and-switch (advertising a product at a price with no or extremely limited inventory, then pressuring consumers toward a more expensive alternative).

**Drip pricing specifically**: The FTC issued a policy statement in 2022 and a proposed rule in 2023 (the "Junk Fees" rulemaking) targeting any practice of advertising a price that does not include mandatory fees. Under existing Section 5 analysis, a hotel that advertises "$89/night" and adds a $35/night resort fee is already potentially deceptive. The proposed rule would make this a penalty-eligible violation. FTC v. IronNet Cybersecurity (2024) and several hotel enforcement actions are recent precedents.

**Enforcement pattern**: Pricing deception enforcement often follows consumer complaints aggregated through the Consumer Sentinel database. The FTC also conducts sweeps (Operation Sticker Shock-type enforcement) of specific industries with a history of pricing deception (car dealers, ticket sellers, hotels).

**What compliant looks like**:
- Reference prices ("Was [X]") reflect a real price at which the product was sold in the recent past (within the prior 90 days at the advertising location is a common standard; the FTC and several state laws require this)
- All mandatory fees are included in the advertised price or prominently disclosed adjacent to the price before the consumer selects the product — not revealed at checkout
- Limited-time and limited-quantity claims are true; if the offer renews daily, "today only" is deceptive

---

## 4. Negative option billing and subscription traps

**The violation pattern**: A consumer signs up for a "free trial" or introductory offer. Buried in the enrollment flow — in fine print, pre-checked boxes, or on a subsequent screen — are terms that convert the offer to a paid subscription unless the consumer affirmatively cancels. Cancellation requires calling a number, sending a physical letter, or navigating a multi-step online process designed to cause drop-off.

**The specific FTC trigger**: Under the 2023 Negative Option Rule (16 CFR Part 425), three requirements apply before billing can begin: (1) clear and conspicuous disclosure of all material terms before billing information is collected, (2) separate affirmative consent to the negative option feature, and (3) a cancellation mechanism at least as simple as enrollment. All three must be present; each failure is independently actionable.

**Enforcement pattern**: Operation Dark Patterns (2022 FTC sweep) produced enforcement actions against ABCmouse, Credit Karma, Vonage, and others for subscription trap practices. The consent orders required simple cancellation flows and disgorgement. The 2023 Negative Option Rule adds direct civil penalty exposure for first violations (not just post-cease-and-desist).

**What compliant looks like**:
- Trial enrollment page states, adjacent to the CTA: the trial length, the exact price that will be charged when the trial ends, the billing date, and how to cancel — all before the consumer enters payment information
- A separate checkbox (unchecked by default) for the subscription/auto-renewal feature — distinct from the terms of service checkbox
- Cancellation available through the same channel as enrollment (enrolled online → cancel online); no phone-only or email-only cancel for online sign-ups
- Cancellation confirmation sent to the consumer; no "we'll get back to you in 5-7 business days to confirm your cancellation"

---

## 5. Fake reviews and manipulated social proof

**The violation pattern**: A brand creates, purchases, or incentivizes reviews that do not reflect genuine consumer experience — AI-generated reviews, reviews from company insiders without disclosure, paid five-star reviews from review farms, review suppression (publishing only positive reviews from a solicited set while ignoring negative ones), or review hijacking (repurposing reviews from a different product or time period).

**The FTC's 2024 Fake Reviews Rule explicitly covers**:
- Buying reviews from any source, including AI-generated content
- Company insider reviews without prominent disclosure of the relationship
- Reviews solicited through a mechanism that filters or discourages negative responses (e.g., "if you're happy, click here to leave a review" without an equivalent path for unhappy customers)
- Repurposing reviews from a different product, formulation, or seller as current reviews for the advertised product
- Using company-controlled social media personas to post reviews

**Enforcement pattern**: The FTC issued a record $26.5 million civil penalty against Lowe's for pay-per-review practices (2024). Operation AI Comply produced enforcement actions for AI-generated fake reviews. Review manipulation is also aggressively enforced by the FTC's Consumer Sentinel network and coordinated state AG actions.

**What compliant looks like**:
- Review solicitation sends equal prompts to all purchasers (not just those who indicated satisfaction in a post-purchase survey)
- All insider or employee reviews are labeled with the relationship ("Employee review" or equivalent)
- No financial or non-financial incentives are offered for reviews without requiring disclosure of the incentive in the review itself
- AI tools are not used to generate content that is then published as consumer review text

---

## 6. Misleading before/after images

**The violation pattern**: A product's advertising uses before/after images (weight loss, skin care, hair restoration, home improvement, teeth whitening) that are created or altered in a way that does not represent genuine, achievable results from using the product as directed — including use of lighting, makeup, posing, image editing, AI image generation, or models who did not actually use the product.

**Why this is an unsubstantiated claim**: A before/after image is a visual claim that the depicted result is a real, typical result from product use. The FTC treats it as an objective performance claim subject to the same substantiation standard as a written claim. The asterisk "results not typical" was explicitly addressed in the 2023 Endorsement Guides: it does not cure a deceptive before/after if the image shows atypical results; instead, the image itself must depict typical results or must clearly disclose what the depicted result actually reflects.

**Enforcement pattern**: Weight loss and cosmetic product before/after claims have been FTC enforcement targets for decades. AI-generated before/after images were specifically addressed in Operation AI Comply (2024) — one of the five actions involved AI-generated images of skin improvement that did not reflect real product results.

**What compliant looks like**:
- Before/after images reflect real results from real consumers who used only the advertised product as directed (not combined with diet, exercise, other products, or other treatments that the advertiser does not also sell)
- If results shown are above average, the image is accompanied by a disclosure of what a typical user can expect ("In a study of 50 users, average weight loss was X lbs; this user lost Y lbs")
- No post-production alterations that affect the appearance of the "result" (cropping for posture change, lighting adjustments on "after" that were not present on "before")
- AI-generated before/after images are not used to represent product results

---

## 7. Native advertising without adequate disclosure

**The violation pattern**: An advertiser pays for content — an article, a video, a social post, a podcast episode segment, a search placement — that is designed to look like editorial, journalistic, or organic content. The paid nature of the placement is either not disclosed or disclosed in a way that is not clear and conspicuous (e.g., "sponsored" in light gray 8pt text, or disclosed only in metadata that readers never see).

**The FTC's "Native Advertising" guidance** (2015, still operative): Paid content that is not clearly distinguishable from editorial content is deceptive if consumers are likely to believe it is independent editorial, and if the editorial nature (the implied independence) is material to how consumers evaluate the content. The disclosure must be presented so that consumers notice it, read it, and understand it before engaging with the content — not after.

**Enforcement pattern**: The FTC issued its first native advertising enforcement action against Lord & Taylor in 2016 (paid magazine placements and influencer posts presented as editorial). Since then, enforcement has extended to sponsored search result formatting, branded podcasts, and "content marketing" articles. Publisher liability is also on the table when publishers design ad placements to be indistinguishable from editorial.

**What compliant looks like**:
- Sponsored content is labeled with clear language — "Sponsored," "Advertisement," "Paid Content by [Brand]" — in a font size and color that contrasts with the background and is visible before the reader engages with the body content
- The disclosure label is placed before (above) the headline or at the top of the content — not at the bottom, in metadata, or in a publication's general advertising disclosure page
- "Partner content," "presented by," "brought to you by" without the word "sponsored," "advertisement," or "paid" are insufficient — the FTC has specifically flagged these as ambiguous
- Sponsored search listings use the publisher's required disclosure format (Google's "Sponsored" label, etc.) and the advertiser does not request or encourage formatting that obscures the paid placement
<!--fold:b41618@file path="sources.md" mode="644"-->
# FTC Marketing Compliance Sources

Penalty amounts, enforcement priorities, and guidance documents change. Use these URLs to pull current versions. This file maps the structure; the live URLs are the authoritative source.

---

## FTC primary guidance

**FTC .com Disclosures Guide**
- https://www.ftc.gov/business-guidance/resources/disclosures-how-make-effective-disclosures-digital-advertising
- The operative guide for digital advertising disclosures: what "clear and conspicuous" means in digital contexts, space-constrained media (social, mobile), hyperlinked disclosures, and platform-specific considerations. Start here for any question about disclosure adequacy.

**FTC Endorsement Guides (2023 revision)**
- Guide text and FAQ: https://www.ftc.gov/business-guidance/resources/ftcs-endorsement-guides-what-people-are-asking
- Final rule codified at 16 CFR Part 255: https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-255
- Covers influencer marketing, affiliate marketing, employee advocacy, review solicitation, virtual influencers, and social media endorsements. The 2023 revision added explicit treatment of non-cash material connections, virtual influencers (AI personas), and the "clear and conspicuous" standard by platform type.

**FTC Negative Option Rule (2023 final rule)**
- Business guidance: https://www.ftc.gov/business-guidance/resources/negative-options-guide-businesses
- Final rule codified at 16 CFR Part 425: https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-425
- Covers auto-renewal subscriptions, free-to-paid trial conversions, continuity plans, and any offer where silence equals consent. The 2023 rule added the click-to-cancel (simple cancellation) requirement and mandatory affirmative consent.

**FTC Fake Reviews and Testimonials Final Rule (2024)**
- Rule text: https://www.ftc.gov/legal-library/browse/rules/trade-regulation-rule-use-consumer-reviews-and-testimonials
- Prohibits buying or creating fake reviews, AI-generated fake reviews, insider reviews without disclosure, review suppression, and review hijacking. Civil penalties per violation. Effective October 2024.

**FTC Operation AI Comply (2024)**
- Press release and enforcement action links: https://www.ftc.gov/news-events/news/press-releases/2024/09/ftc-announces-crackdown-ai-enabled-deception-launches-operation-ai-comply
- Five enforcement actions against companies using AI to generate fake reviews, fake testimonials, and deceptive weight-loss claims. Clarifies that Section 5 deception liability applies regardless of AI origin of content.

**FTC CAN-SPAM Compliance Guide**
- https://www.ftc.gov/business-guidance/resources/can-spam-act-compliance-guide-business
- Covers all seven CAN-SPAM requirements, the 10-business-day opt-out window, the "primary purpose" test for mixed commercial/non-commercial messages, physical address requirement, and third-party sender liability.

**FTC AI and Algorithmic Guidance**
- FTC AI policy page: https://www.ftc.gov/policy/advocacy-research/tech-at-ftc/ai
- Business guidance blog posts on AI, deception, and Section 5: https://www.ftc.gov/business-guidance/blog
- The FTC publishes AI-specific guidance as blog posts and policy statements rather than in a single document; search the business guidance blog for "artificial intelligence" for current enforcement framing.

---

## FTC enforcement database

**FTC Cases and Proceedings (searchable)**
- https://www.ftc.gov/legal-library/browse/cases-proceedings
- All FTC enforcement actions, consent decrees, and court orders. Filter by topic (advertising, endorsements, negative option, COPPA, AI) to find current enforcement posture. Check recent cases (past 18 months) in the relevant vertical before advising on compliance exposure.

---

## Email and electronic message compliance

**CASL (Canada) — statute and CRTC guidance**
- CASL statute: https://laws-lois.justice.gc.ca/eng/acts/E-1.6/
- CRTC compliance and enforcement: https://crtc.gc.ca/eng/internet/anti.htm
- CRTC guidance covers implied vs. express consent, the 24-month purchase window, the 6-month inquiry window, and enforcement procedures. The CRTC, the Competition Bureau, and the Privacy Commissioner share CASL enforcement jurisdiction.

**TCPA (US SMS / telephone)**
- FCC TCPA resources: https://www.fcc.gov/consumers/guides/stopping-unwanted-robocalls-and-texts
- SMS marketing requires prior express written consent for autodialed or prerecorded messages; separate from CAN-SPAM; the FCC enforces TCPA. Consult TCPA counsel for SMS campaign design.

---

## Children's advertising

**COPPA Rule and Business Guidance**
- Business FAQ: https://www.ftc.gov/business-guidance/resources/complying-coppa-frequently-asked-questions
- Full COPPA rule at 16 CFR Part 312: https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-312
- Verifiable parental consent methods, what constitutes a site "directed to children," mixed-audience site rules, and data retention/deletion requirements. The FTC is actively reviewing COPPA for potential age threshold updates.

---

## State consumer protection laws

**California**
- Unfair Competition Law (UCL): Cal. Bus. & Prof. Code §§ 17200–17210
- False Advertising Law (FAL): Cal. Bus. & Prof. Code §§ 17500–17509
- Consumer Legal Remedies Act (CLRA): Cal. Civ. Code §§ 1750–1784
- California AG enforcement: https://oag.ca.gov/consumers
- California Privacy Protection Agency (CPPA) — AI and automated decision-making rules in development: https://cppa.ca.gov/regulations/

**Multi-state UDAP reference**
- National Consumer Law Center state consumer protection survey: https://library.nclc.org/udap/state-summaries
- Each state's UDAP statute, whether it allows private rights of action, and penalty structure.

---

## How to use these sources

1. **Influencer or affiliate campaign** → Endorsement Guides (2023) first, then .com Disclosures guide for platform-specific format questions
2. **Substantiation question** → FTC Advertising FAQ (linked from the .com Disclosures guide) and the endorsement guide section on claim types
3. **Subscription or free trial flow** → Negative Option Rule final rule text (not just the business guidance summary); check for state auto-renewal laws (California, New York, and others are stricter)
4. **Email campaign with Canadian recipients** → CASL statute and CRTC guidance; do not apply CAN-SPAM analysis to Canadian recipients
5. **SMS campaign** → TCPA (FCC), CASL if Canadian recipients, and state texting laws; consent requirements are stricter than email
6. **AI-generated content** → Operation AI Comply enforcement actions for current FTC posture; Fake Reviews Final Rule for review-specific conduct; state synthetic media disclosure laws for political and some commercial contexts
7. **Children's product or service** → COPPA rule (16 CFR Part 312) and FTC FAQ; California CPPA rules for children's data
8. **Recent enforcement posture** → FTC Cases and Proceedings, filtered to the relevant topic, sorted by date; the FTC's stated priorities and actual enforcement actions sometimes differ
<!--fold:b41618@end-->
PORTDOWN_A0CBD61B

# ── post ──
MARKER=$(awk '/^---$/ { f++; if (f==2) exit; next } f==1 && /^marker:[[:space:]]/ { sub(/^marker:[[:space:]]+/, ""); print; exit }' "$DEST")
[ -z "$MARKER" ] && { echo "seed: archive has no marker — corrupt" >&2; exit 1; }
awk -v m="$MARKER" -v outdir="$TARGET" '
  BEGIN {
    # Match <!--fold:<m>@file path="X"--> with an optional mode attr after
    # the path (fold emits  mode="644"  on executables).
    file_re = "^<!--fold:" m "@file path=\"([^\"]+)\"( mode=\"[0-9]+\")?-->$"
    end_re  = "^<!--fold:" m "@end-->$"
  }
  $0 ~ end_re { if (current) close(current); exit }
  $0 ~ file_re {
    if (current) close(current)
    line = $0
    sub(/^<!--fold:[^@]+@file path="/, "", line); sub(/".*$/, "", line)
    current = outdir "/" line
    dir = current; sub(/\/[^\/]*$/, "", dir)
    if (dir != current) system("mkdir -p \"" dir "\"")
    printf "" > current
    next
  }
  current { print >> current }
' "$DEST"
SEED_EXTRACTED=$(find "$TARGET" -type f -not -path "$DEST" 2>/dev/null | wc -l)
if [ "$SEED_EXTRACTED" = "0" ]; then
  echo "seed: archive contained no files — refusing to delete the source" >&2
  echo "  archive preserved at: $DEST" >&2
  exit 1
fi
rm -f "$DEST"

echo "" >&2
echo "✓ seed unpacked → $TARGET ($SEED_EXTRACTED files)" >&2
find "$TARGET" -type f | sort | while IFS= read -r _sf; do
  echo "  ${_sf#${TARGET}/}" >&2
done
echo "" >&2
if [ -f "$TARGET/SKILL.md" ]; then
  echo "This seed contains a skill (SKILL.md). Install it in your agent's skills directory." >&2
  echo "" >&2
fi
echo "Install the seed skill if not already installed:" >&2
echo "  https://seed.show/skill" >&2
echo "" >&2
echo "Publisher prompt:" >&2
sed 's/^/  /' >&2 <<'__SEED_PROMPT_END_AC1F2B__'
You are an AI agent helping review marketing copy, campaigns, or advertising strategy for FTC compliance. README.md gives you the structural framework — two-pronged FTC authority, the deceptive practices standard, and what agents most commonly get wrong (influencer disclosures, claim substantiation, negative option flows, AI-generated content, fake reviews, email jurisdiction). failure-modes.md gives you the seven most common failure patterns and what compliant practice looks like for each. sources.md maps the primary FTC guidance documents and enforcement databases you should fetch for current penalty amounts and enforcement posture. Start with README.md, then pull sources.md before any live compliance review. If the question involves actual enforcement exposure or specific penalty risk, say so and recommend legal counsel.
__SEED_PROMPT_END_AC1F2B__
exit 0