EXHIBIT 01 RISK ASSESSMENT — REGULATED HEALTH ADVERTISING

Your ad is one screenshot away
from a warning letter.

Every claim you publish about a GLP-1, a peptide, or a "results in days" protocol is a liability record. ClaimRight scans it before it goes live, rewrites what's flagged, regenerates the creative, and re-checks the rewrite until it actually passes — across every platform you advertise on, at once.

FILE NO. CR-2026-0417STATUS: UNDER REVIEW
FLAGGED — DO NOT PUBLISH
38/ 100 — sample scan
Red blocker phrases3 found
Yellow flag patterns2 found
Platforms at riskMeta, Google, TikTok
Evidence on fileInsufficient
SAMPLE DATA — illustrative scan output, not a real client file
Science applied in this section
Loss Aversion (Kahneman) Von Restorff Effect Specificity Effect
EXHIBIT 02 THE ENFORCEMENT REALITY

Regulators are not guessing. They're pattern-matching — and so are the platforms.

The FTC has an active, ongoing enforcement posture against deceptive health and weight-loss advertising, with a well-documented history of actions against unsupported efficacy claims, fake before/after results, and manipulated testimonials. The FDA separately polices claims that position a product as treating, curing, or preventing disease — a line that peptide, GLP-1, and longevity advertisers cross constantly, often without realizing the specific phrase that tripped it.

Layered on top of that: Meta, Google, and TikTok all run their own automated ad-policy classifiers for health and pharmaceutical content. Your ad doesn't need a federal investigation to die — it just needs an algorithm to flag "drug-adjacent" language, and the account-level consequences (rejected ads, suspended ad accounts, shadow-restricted reach) compound fast.

WHERE ADS ACTUALLY DIE
Failure pointWhat triggers it
Regulatory letterUnqualified outcome claims ("cures," "eliminates," "guaranteed")
Platform rejectionDrug-adjacent language, before/after imagery, fake credentials
Account suspensionRepeated policy strikes across campaigns
Silent reach throttleClassifier flags without a formal rejection notice

Most clinics find out which rule they broke only after the ad account is already restricted — the fix has to happen before the click "Publish," not after.

Science applied in this section
Loss Aversion (Kahneman) Authority (Cialdini) Specificity Effect
EXHIBIT 03 SIX MODULES, ONE WORKBENCH

Scan → Rescue → Generate → Publish-safe → Test → Learn. On loop.

ClaimRight is not a checklist. It's an AI compliance workbench with six connected modules, each doing real work — not a single scanner bolted onto a copy generator.

MODULE 01 — SCAN

Live rule-engine scoring

Paste or upload ad copy. Scored against 12 red-blocker phrase patterns and 8 yellow-flag patterns, checked across Meta, Google, TikTok, SMS, Email, and Landing Page simultaneously. Output: score/100, tier chip, blocked phrases, yellow flags, evidence-needed notes, per-platform risk.

MODULE 02 — RESCUE

AI rewrite, verified by re-scan

Every flagged ad gets an AI-generated before/after rewrite. The rewrite is automatically re-scanned — it has to actually pass, not just claim to. No rewrite ships on a promise.

MODULE 03 — IMAGE

Brand-consistent regeneration

Generates an image-regeneration prompt for the rescued ad with baked-in negative constraints: no fake member counts, no fake badges, no before/after body imagery, no needle imagery. Local preview render included.

MODULE 04 — SOCIAL

Six platform-safe variants

Meta-safe, Google-safe, TikTok-safe (with #sponsored disclosure), SMS-safe (STOP opt-out included), Email-safe (CAN-SPAM compliant), Landing-safe. Each with caption, hashtags, and character-count check per platform. Every output flagged "human review mandatory."

MODULE 05 — PERSONAS / TEST

Audience testing on real data

Archetype testing seeded from your actual uploaded analytics — the derived stats change persona weights, they're not cosmetic. A junk-data detection gate blocks obviously fake or manipulated upload data before it can poison the test.

MODULE 06 — FUNNEL

7-stage funnel, one style-lock

A 7-stage funnel generator with a style-lock — character, phrase hook, palette, font pair — carried consistently across every stage so the funnel doesn't fragment into six different brand voices.

MODULE 07 — BULK

Queue and run at scale

Queue multiple ads, scan and rescue at scale, status pills per ad, run-all execution — for clinics running more than one campaign at a time.

MODULE 08 — LEDGER

Full audit trail, exportable

Every scan, rescue, and publish action logged in a state machine per ad: draft → blocked → needs_evidence → approved_internal → approved_platform → live. Exportable as CSV or JSON for your compliance file.

Watch what one of these scans actually looks like on a real sample ad — scroll to the live terminal below.

Science applied in this section
Specificity Effect Commitment Ladder Foot-in-the-door
EXHIBIT 04 LIVE SCAN — SAMPLE OUTPUT

What the scan terminal actually shows you

Illustrative sample data — not a real client scan
claimright — scan_engine v1 — session #A417
> scan --input "ad_copy_GLP1_launch.txt" --platforms all
Loading rule engine: 12 red_blockers, 8 yellow_patterns, 6 platform rulesets...
Parsing 214 words...
RED — blocker phrase: "guaranteed weight loss" (unqualified outcome claim)
RED — blocker phrase: "cures your metabolism" (disease-treatment claim, FDA scope)
RED — blocker phrase: "melt fat in days" (unqualified rapid-outcome claim)
YELLOW — flag: testimonial lacks "results not typical" disclosure
YELLOW — flag: no compounding/individual-results disclaimer present
Evidence check: clinical citation — NOT ON FILE
> score: 38 / 100 — tier: BLOCKED
Meta
High risk
Google
High risk
TikTok
High risk
SMS
Moderate
Email
Moderate
Landing
Low risk

That's the same scan that runs before every rewrite, every image regeneration, and every platform-safe variant below — nothing ships without passing this gate first.

Science applied in this section
Von Restorff Effect Page Scent Specificity Effect
EXHIBIT 05 THE METHODOLOGY BEHIND THE OUTPUTS

Every rewrite ClaimRight generates is built on the same behavioral-design principles it's checking your ad against.

01

Loss Aversion

Kahneman & Tversky. Framing a warning letter or account suspension as an already-in-motion risk drives faster review than framing "more conversions" as an upside.

02

Specificity Effect

Concrete, named numbers and mechanisms ("12 red-blocker patterns, 6 platforms") read as more credible than vague claims — the same logic your rescued copy applies to your offer.

03

Anchoring

Every price comparison below anchors against real consultant and agency market rates before ClaimRight's own quote-only positioning is introduced.

04

Commitment Ladder

Cold ask (see the scan) → warm ask (scan a sample ad) → hot ask (book the workbench demo). No page asks for a contract on the first click.

What compliance actually costs you right now — without an AI layer

These are real, publicly-cited market rates for the manual version of what ClaimRight automates. ClaimRight itself is quote-only — no public price is listed anywhere on this page.

Provider typeWhat you getMarket rate
Compliance consultantOne manual audit, no rewrite, no re-check, no AI$5,000 – $15,000 per audit
Ad agency retainerManual creative production, zero compliance layer$10,000+ / month
CRO consultantManual test design and analysis, hourly$200+ / hour
ClaimRightScan + rescue + image + social + personas + funnel + analytics, continuous, one systemQuote-only, by license

One integrated AI system, continuously running the scan/rescue/generate/test/analyze loop, versus a stack of hourly consultants who each touch one piece of it once.

Science applied in this section
Anchoring Decoy Effect Authority + Unity (Cialdini)
EXHIBIT 06 RESCUE MECHANISM — BEFORE / AFTER

The rewrite has to earn its pass. It doesn't get to claim it.

Sample ad copy — illustrative, not a real client file
▲ BEFORE — BLOCKED, score 38/100
"Our clinic's guaranteed weight loss protocol melts fat in days — patients report it cures their metabolism for good."
Rule fired: red_blocker #03 (unqualified outcome claim), red_blocker #07 (disease-treatment claim), red_blocker #11 (unqualified rapid-result claim)
● AFTER — RE-SCANNED, score 91/100
"Our clinic's physician-supervised weight-management protocol is designed to support your metabolic goals. Individual results vary — ask our team whether this program is right for you."
Why it passes: outcome claim qualified, disease-treatment language removed, results-vary disclosure added — re-scanned automatically, not asserted.

This is the exact mechanism running on every ad in your queue — flagged, rewritten, and re-verified before it ever reaches a platform.

Science applied in this section
Specificity Effect Von Restorff Effect Authority (Cialdini)
EXHIBIT 07 THE ANALYTICS LOOP — YOU STAY IN CONTROL

Not a black box. A loop you approve at every stage.

ClaimRight doesn't silently decide what runs against your audience. It proposes, you approve, and the system tracks what actually reduced risk or improved performance — not what it merely predicted would.

ANALYZE
HYPOTHESIS
PREDICTION
YOU ACT
SELF-EVALUATION
STEP 1-2

Analyze + Hypothesize

The system parses your uploaded analytics — real row counts, real date ranges, real device/source columns — and proposes a specific test hypothesis, not a generic one.

STEP 3-4

Predict + You Act

A prediction is logged before the test runs. You choose which variant launches — the system never auto-publishes a test without your approval.

STEP 5

Self-Evaluation

Results are checked against the original prediction. What worked gets promoted to the funnel style-lock; what didn't gets logged to the ledger, not repeated blindly.

The junk-data gate matters here too — a manipulated analytics upload gets blocked before it can poison a single persona weight.

Science applied in this section
Commitment Ladder Foot-in-the-door Authority (Cialdini)
EXHIBIT 08 WHO OPENS THIS FILE

Built for operators who advertise regulated outcomes, not commodity products.

A

Telehealth operators

Multi-state ad accounts, multi-platform spend, and a compliance surface that grows with every new state license.

B

GLP-1 / weight-loss clinics

The single highest-scrutiny category in health advertising right now — outcome claims are the first thing regulators and platforms both flag.

C

Peptide / longevity clinics

Compounded, off-label, and research-use language all carry distinct risk profiles that a generic compliance checklist won't separate.

D

Medspas

Aesthetic outcome claims plus before/after imagery — two of the most commonly flagged ad elements across every platform.

If your ad account has ever been rejected, restricted, or throttled without a clear explanation — the next section answers the question you're actually asking.

Science applied in this section
Specificity Effect Von Restorff Effect
EXHIBIT 09 OBJECTIONS ON THE RECORD

The questions every serious buyer asks before signing.

Q. Is this legal advice?
No. ClaimRight is a compliance workbench, not a law firm. It flags known risk patterns and generates rewrites; every output requires human review before publishing, and nothing here replaces your compliance attorney's judgment on your specific claims.
Q. What if I already got a warning letter or a platform suspension?
ClaimRight can scan the flagged copy and generate a rescue rewrite immediately, but an active enforcement action needs your attorney in the loop from the start — we'll say that plainly on the call, not bury it.
Q. Does this replace my compliance lawyer?
No. It reduces how often you need one on retainer for routine ad review, and gives your lawyer a cleaner starting point when you do need one.
Q. Do you guarantee no fines or account suspensions?
No tool can guarantee regulatory or platform outcomes. What ClaimRight guarantees is that every rewrite it ships has been re-scanned and passed the same rule engine that flagged the original — before a human ever approves it.
Q. What platforms does it cover?
Meta, Google, TikTok, SMS, Email, and Landing Page — scored simultaneously on every scan, with platform-specific safe variants generated for each.
Q. Can it read an image ad, not just text?
You can upload an image alongside your copy for reference; the scan itself runs on the ad text and claims. Image-specific risk (before/after body imagery, fake badges) is checked as part of the rescue and regeneration step.
Q. What data do you need from me?
Your ad copy to start. For the Personas module, your own analytics export (CSV) — parsed to seed real audience weights, gated against obviously fake or manipulated uploads.
Q. How fast is a rescan?
The rule-engine scan itself runs in seconds; the AI rewrite and re-verification loop typically completes well within a single working session, not days.

If your questions aren't fully answered above, the walkthrough call covers your exact ad account and claim history.

Science applied in this section
Authority (Cialdini) Specificity Effect
EXHIBIT 10 REQUEST FILING

Request a private quote — no public pricing, no generic package.

Every ClaimRight license is quoted against your ad volume, platform mix, and current compliance exposure. Tell us where you are and we'll walk you through the workbench live on your own ad copy.

STATIC FORM UI — NOT WIRED TO A BACKEND
This is a static file — the form fields above are UI mockups only (disabled). Use the mailto link to actually reach us.

What happens after you request a quote

  1. We review one of your live ads against the full rule engine and send back the real scan output — not a demo.
  2. We walk through the rescue, image regeneration, and platform-safe variants live on the workbench.
  3. You get a quote scoped to your ad volume and platform mix — no public tier to negotiate against.

This is the last click on the page — every question above this point has been answered before you're asked to reach out.

Science applied in this section
Scarcity Commitment Ladder Fitts's Law (CTA sizing)