Insurance Advertising Compliance Checker

Insurance advertising regulations, and who writes them.

Insurance ads answer to the states, and most states build their rules from the same handful of NAIC model laws. Here is what those models say about an ad, what changes by line, where Medicare adds a federal layer, and where a producer's own review sits before the carrier's.

What insurance advertising compliance is

Insurance is regulated by the states, and so is insurance advertising. There is no single federal rulebook for an agent's Facebook post or an agency's landing page. Instead, each state's insurance department enforces its own insurance advertising laws, and most of those laws are built on model acts and regulations written by the National Association of Insurance Commissioners (NAIC). A model is a template. A state adopts it, often with changes, and the version your state adopted is the one that binds you.

Two models do most of the work. The Unfair Trade Practices Act (Model #880) applies to every line: it defines misrepresenting "the benefits, advantages, conditions, or terms of any policy" as an unfair trade practice, and it reaches any advertisement or statement about the business of insurance that is "untrue, deceptive or misleading", explicitly including "electronic mail, internet advertisement or posting". Then each line has its own advertising regulation: Model #570 for life insurance and annuities, and Model #40 for accident and sickness insurance. Medicare plans add federal rules from CMS on top of all of it.

The rules, by line

Each pack in this checker cites the models and regulations it was built from. The main ones, with what each governs:

NAIC Model 570: the life and annuity ad rules

Model #570 starts from one test. Its Section 4A says advertisements "shall be truthful and not misleading in fact or by implication", and that whether an ad can mislead is judged "from the overall impression that the advertisement may be reasonably expected to create upon a person of average education or intelligence within the segment of the public to which it is directed." That is the average-consumer test, and it is why a technically accurate sentence can still be a problem: the question is what the reader takes away.

From there the model names specific traps. Section 4B bars terms such as "investment", "profit", "savings", "savings plan" and "retirement plan" where they could lead a buyer to think they will get something other than a life policy. Section 5J says analogies between cash values and savings accounts "shall be complete and accurate", and that an ad "shall not emphasize the investment or tax features" so far that a buyer believes the policy "is anything other than life insurance." Dividends may not be called "tax-free" unless the tax treatment is fully explained and the dividend is shown to be a return of premium.

Two rules matter most for simplified-issue and guaranteed-issue ads. Under Section 5C, "non-medical" or "no medical examination required", used where issue is not guaranteed, must sit beside a disclosure of equal prominence that issue may depend on the answers to the health questions. And under Section 5H, the graded-benefit and similar policies it covers may not be advertised with the words "inexpensive" or "low cost" when they are guaranteed issue. Section 6A then requires that "the name of the insurer shall be clearly identified in all advertisements about the insurer or its products".

Health insurance: Model 40, and Medicare's federal layer

Model #40 applies the same overall-impression test to accident and sickness ads and adds its own requirements: an ad that invites a purchase must disclose the exceptions, reductions and limitations affecting the policy's basic provisions, a drafting note names phrases that exaggerate a benefit's effect on the buyer's well-being, such as "worry-free savings plan" and "financial peace of mind" (the life pack flags those phrases; the health pack does not today), testimonials "shall be genuine, represent the current opinion of the author" and be reconfirmed if used more than a year after they were given, and under Section 14 "the name of the actual insurer shall be stated in all of its advertisements."

Medicare Advantage marketing answers to CMS as well. Under 42 CFR 422.2262, a plan may not "claim they are recommended or endorsed by CMS, Medicare, the Secretary, or HHS", may not use the term "free" to describe a $0 premium or a reduction in cost sharing, and may not use the Medicare name or the Medicare card "in a misleading way." The same section allows a plan to say it is approved to participate in Medicare and to use "Medicare-approved" to describe benefits or services, which is why a phrase checker has to be careful with the word. An agent selling for more than one plan also carries the third-party marketing disclaimer in 42 CFR 422.2267(e)(41), which begins "We do not offer every plan available in your area." This checker does not test for that disclaimer on its own; it looks only for the health pack's general disclaimer.

Property and casualty: coverage, claims and savings

Property and casualty has no line-specific advertising model like #570 or #40 among the rules these packs cite. As the P&C pack describes it, auto, home and commercial ads are policed through Model #880's misrepresentation rule and each state's claims-practice statute, mostly after the fact rather than by pre-filing. The P&C pack encodes the claims that turn up in that enforcement: coverage broader than any policy provides ("covers everything", "no exclusions"), a claims process that sounds automatic ("every claim is paid"), a specific savings figure stated as assured ("save $500"), and a comparison with competitors with no stated basis. "Full coverage" is a warning, not a hard flag, because the phrase has no fixed meaning and is only a problem where it overstates scope. The pack cites Washington's and New York's savings rules for that last category; we paraphrase them here because we did not fetch those statutes for this page.

The agent's side: carrier review comes after yours

Most writing about insurance advertising is aimed at carrier compliance teams. The models put the agent in the frame too. Model #570, Section 3B makes every advertisement "the responsibility of the insurer, as well as the producer who created or presented the advertisement", and requires insurers to notify agents at least once a year "of the requirement and procedures for company approval prior to the use of any advertisements that is not furnished by the insurer." In practice that is the carrier's advertising review: an ad you wrote that names a carrier or its products goes to their desk before it runs.

That review is where this tool fits, and where it stops. It is a pre-check an agent or an agency runs before the ad reaches the carrier, so the obvious phrases are gone before a reviewer spends time on them. It is not a substitute for the carrier's sign-off, for your agent agreement's own standards, or for your state's version of the rules.

A phrase library, by line

These are real rules from the packs this checker runs, each with a phrase it flags. They are examples of the pattern, not the whole pack, and every one was tested against the rule it sits beside. The three line packs are drafts, with coverage still being built.

Why the required disclaimer is checked by phrase

Each line pack carries one general disclaimer block, and a page is expected to carry it once. Agencies word these blocks their own way, so the checker does not demand the exact text. It looks for the phrases that carry the weight: for the life pack, "informational", "does not constitute insurance", "guaranteed beyond the policy", "tax" and "consult a licensed". The health block adds "not affiliated with", for the government and the Marketplace; the P&C block adds "exclusions". If any is absent, the disclaimer is flagged with a count of how many.

Below is a real run. It is five paragraphs from an article on verand.ai for life insurance agents, about whether software may draft their marketing, pasted as published and checked against the life line. It makes no promise about a policy, so no banned claim matched. The disclaimer was flagged because the article carries none of the life block's phrases: it is an article about advertising, not an ad for a policy. That flag may matter or may not, which is exactly the call a flag leaves to a person.

Manual review versus a phrase checker

A compliance reviewer reads for the overall impression Model #570 asks about: who the audience is, whether the insurer is named, whether a premium is described fairly, whether a testimonial is genuine and current. None of that is a pattern. What a deterministic checker does well is the other half: it reads every occurrence of every rule the same way every time, never skims the fourth paragraph, and is fast enough to run on every draft rather than only on the version that goes to the carrier.

The price of that consistency is literalness. "Every policy comes with a guaranteed payout" is flagged; "the policy pays out whatever happens" makes the same promise in words no rule matches. On the life line, "No medical exam required" is flagged for review unless a phrase such as "may depend" or "health questions" sits beside it, and even then a phrase check cannot tell whether that disclosure has the equal prominence Model #570 asks for. Treat a clean result as the absence of the listed problems, not as a green light.

Common mistakes

• Treating a clean scan as sign-off. No tool, this one included, can tell you an ad is allowed. It can tell you it did not find the patterns it knows.
• Skipping the carrier because the scan was clean. An ad that names a carrier or its product still goes through the carrier's review, on the terms of your agent agreement.
• Checking the headline, not the page. Disclaimers often sit in a footer or a template. Paste the whole rendered page, or the disclaimer will read as missing.
• Picking the wrong line. A Medicare page checked against the life line misses every Medicare rule. Pick each line the page actually sells; the first one supplies the disclaimer.
• Qualifying a claim somewhere else. "Guaranteed issue" in the fine print does not qualify "guaranteed acceptance" in the headline. The exception has to sit beside the phrase.
• Forgetting who reviewed the rules. The line packs are drafts, with coverage still being built. Weigh their flags, and their silences, accordingly.