ppl.studio

AI UGC for Insurance and Financial Services Marketing: The Image Is Part of the Claim

Insurance agents, advisors, banks and fintechs can use AI UGC for the life-moment imagery these products are sold on: a first home, a new baby, a teenager's first car, a retirement trip. It works for Facebook and LinkedIn ads, email headers and social posts. The catch is that in a regulated ad the image is part of the claim. A picture can imply coverage you don't offer, a client who doesn't exist, or an outcome you can't promise, and your compliance reviewer will read it that way.

4 min read

A couple at a kitchen table reviewing paperwork, an example of AI-generated life-moment imagery

Financial products have no physical form, so the marketing shows the moment the product protects instead. That has always been done with stock photography, and the industry is one of the heaviest users of the same smiling-couple-on-a-couch images. Generated imagery lets an agency or advisor show their own market: a farmhouse rather than a city loft, a multigenerational household, a small-business owner in a real trade. It doesn't change the review rules your marketing already goes through. Plan for those from the start.


The four ways an image becomes a compliance problem

1. It implies coverage

A homeowners ad showing a flooded living room tells the viewer the policy covers floods. Standard homeowners policies generally don't; flood is usually a separate policy (FEMA's FloodSmart says so plainly). The same goes for an auto ad showing a rideshare driver when personal policies often exclude rideshare use, or a renters ad showing a stolen bike that is over the sub-limit. When you generate a scene, ask what loss it depicts and whether the advertised product pays for it. State insurance advertising rules, many based on NAIC model regulations, prohibit ads that misrepresent policy benefits, and an image can do that as easily as a headline.

2. It invents a client

A generated “customer” with a quote (“They saved my family after the fire”) is a testimonial from someone who doesn't exist. The FTC's rule on consumer reviews and testimonials (16 CFR Part 465) prohibits testimonials that misrepresent who gave them, and names AI-generated ones. For registered investment advisers, the SEC Marketing Rule allows real testimonials only with specific disclosures, and a fabricated one is an untrue statement in an advertisement. Generated people can appear in scenes. They cannot speak as your clients.

3. It implies an outcome

A retiree on a yacht next to your annuity, or a twenty-something on a beach next to an investing app, is an implied performance claim. For broker-dealers, FINRA Rule 2210 requires communications to be fair and balanced and bars promissory or exaggerated content. FINRA's Regulatory Notice 24-09 reminds firms that its rules, 2210 included, apply to content made with generative AI just as they do to anything else. Keep generated lifestyle scenes ordinary: a kitchen table, a driveway, a hospital waiting room. Ordinary scenes are also more believable.

4. It misrepresents who you are

A generated “advisor” or a generated team photo for a one-person agency suggests staff and expertise that aren't there. Use real photos of the licensed people a client will deal with. If you use Animate to turn an advisor's own photo into a short talking-head explainer, treat the result as a communication like any other. It needs the advisor's consent, a reviewed script and a clear note that the video is AI-generated.


Where generated imagery fits by business type

BusinessGood usesKeep real or avoid
Independent insurance agencyLife-event Facebook ads per line (new driver, new home, new baby), renewal-reminder email headers, local seasonal postsAgent and staff photos, claim stories, any loss scene the policy doesn't cover
Carrier or MGARegional variants of national creative, segment tests (renters vs. homeowners), explainer carousels on how a claim worksCustomer testimonials, claim-payout imagery tied to amounts
Financial advisor / RIALinkedIn post imagery, newsletter headers, webinar promos, life-stage scenes (college planning, business sale)Client testimonials without Marketing Rule disclosures, wealth-display imagery, performance charts
Bank or credit unionBranch-community posts, first-account and first-mortgage life moments, fraud-awareness explainersReal member stories without consent, rate claims in the image
Fintech appFeed-native scenes of the app in use (with your real UI screenshots), onboarding explainersMocked UI showing features you don't have, outcome promises
Medicare-focused agentsEducational posts on enrollment periodsAnything not reviewed against CMS marketing rules (42 CFR Part 422, Subpart V)

Medicare and health coverage: extra caution

Medicare Advantage and Part D marketing is governed by CMS rules in 42 CFR Part 422, Subpart V and the matching Part 423 rules, on top of state rules. Plans and their agents have specific requirements for what marketing can say and how it is reviewed. If you sell in this market, check with your FMO or the plan's compliance team before running any new creative, generated or not.


Build it so review is fast

Compliance review is where most financial marketing waits. Generated imagery can shorten it if you make the reviewer's job easy:

  1. Agree on a scene list per product line in advance. For example, auto gets a new driver with a parent, a commute and a parked car at night, with no crashes. Once the scenes are approved as concepts, individual images move faster.
  2. Keep a log. Record the prompt or brief, the date and where each image ran. Broker-dealers already have recordkeeping obligations for communications, and the log makes them easy to meet.
  3. Use the same persona set across campaigns. A small cast built with AI Experts (a young family, a retired couple, a contractor who owns a small business) keeps your ads consistent, so reviewers aren't judging new faces every week.
  4. Keep numbers out of the image. Premiums, rates and returns go in reviewed copy, not baked into a picture where they are easy to miss in an update.
  5. Label where the platform asks. Turn on AI-content labels on TikTok and Meta where they apply. Our guide to AI content disclosure covers the FTC side.

A first campaign for an independent agency

Pick one line, say auto, and one moment: a teenager getting a license. Generate four variants: suburban driveway, rural road, apartment parking lot, and a parent in the passenger seat. Put the agency's real phone number and real agent photo on the landing page. Run the four as separate ad sets to parents in your service area, and send the winner to review as the concept for the next quarter. The Meta ads UGC workflow handles the variant batch. For LinkedIn-first advisors, the same approach works with business-sale and college-planning scenes.


Cost

ppl.studio is $1.99 per week, with 10 free photos to start and no credit card. Images are generated on your own Google Gemini API key and billed by Google: $0.034 per 1K image and $0.050 per 2K image at list price on the default Nano Banana 2.1 model. For a regulated business the tool cost is a rounding error. What matters is that every image goes through the same review as your copy.

This article is general information about marketing practice, not legal or compliance advice. Your firm's compliance function and applicable regulators have the final word.


Life-moment creative your compliance team can approve

Generate on-brand life-event imagery for each product line, then send it through your normal review. No invented clients, no implied coverage.

Start free with ppl.studio

10 free photos · no credit card required

M

Max Zeshut

Founder of ppl.studio. Building AI tools for product marketing teams who need visual content at scale without the production overhead.