The Pinterest Pixel and Health Data: What the FTC's Hims and Hers Complaint Alleges

Yes, Pinterest is named. It sits in paragraph 77 of the FTC's complaint against Hims & Hers Health, Inc., in a list of advertising platforms whose tracking code was allegedly placed on the company's websites. It is also the entry on that list most likely to be skipped by a compliance review, because Pinterest is usually owned by a content or social team rather than performance marketing, and because it does not feel like a tracking platform. It feels like a place to post pictures.

It is not. Pinterest sells conversion outcomes, ships a conversion tag with automatic event collection, offers a server-side Conversions API, and lets advertisers match site visitors back to logged-in Pinterest accounts. All of the machinery a regulator would look at is there. The complaint alleges these technologies transmitted "Events," which paragraph 67 defines as "the actions of website visitors on Hims' website." None of it has been proven. The case was filed on July 29, 2026 in the Northern District of California as Case No. 3:26-cv-7871, and Hims has denied the allegations, says its privacy policy makes clear that users may choose how their data is used, and intends to defend the case.

Curve exists to make this specific problem go away. It is a HIPAA-compliant conversion tracking platform that lets healthcare and telehealth advertisers run and measure Pinterest campaigns without sending protected health information to Pinterest or any other ad platform.

There is a second reason Pinterest deserves its own analysis. The traffic is different. People do not arrive on a telehealth site from Pinterest because they clicked a retargeting ad. They arrive because they were planning something, and the things people plan on Pinterest include losing weight, getting pregnant, clearing their skin, growing their hair back, and quitting drinking.

The Short Version

• Paragraph 77 of the complaint lists Pinterest among the third-party pixels allegedly placed on Hims platforms, alongside Bing, Google Ads, Criteo, Reddit, TikTok, The Trade Desk, StackAdapt and others.
• The alleged violation is not exotic. It is ordinary conversion tracking, running on pages whose subject matter is a medical condition, on a site that publicly promised users it was "100% online, private, and secure" (paragraph 66).
• Pinterest's tag supports automatic event tracking and enhanced match, both of which increase what is transmitted without anyone editing a line of code.
• Pinterest traffic is intent-rich in a way that makes even a bare page view meaningful. A discovery platform tells you what someone is planning, not just what they bought.
• The FTC accurately described Meta's server-side Conversions API in paragraph 70 and pled it as a violation vector anyway. Moving a Pinterest integration to the Conversions API does not change the analysis.
• The remedy is not to stop advertising on Pinterest. It is to control what leaves your servers before it reaches Pinterest.

What the Complaint Actually Says About Pixels

The tracking half of the complaint is compact. Paragraph 66 sets up the promise: Hims published claims that its service was "100% online, private, and secure," that it treated conditions "privately," that the experience was "totally private" and "discreet." The complaint notes those promises appeared in television, radio and podcast advertising as well as on the website.

Paragraph 67 supplies the mechanism, and the definition it uses is the important part. "Events" are "the actions of website visitors on Hims' website." Not form submissions containing a diagnosis. Not medical records. Actions. A page view is an action. Clicking into a treatment category is an action. Starting an intake questionnaire is an action.

Paragraph 70 addresses Meta specifically and names both the Meta Pixel and the Conversions API, describing the latter as operating differently "to the extent it creates a direct connection between the advertiser's server, website, app or other internal software and Meta's systems." The FTC understood the difference and pled both. That single paragraph is why the phrase server-side tracking alone is not HIPAA compliance stopped being an opinion and became the observed position of the enforcing agency.

Paragraph 74 explains why the FTC cares about targeting precision at all: "Hims was only able to create audiences with such specificity because it flouted the promises it made to its users about treating their medical conditions 'privately' or keeping their health information private." Paragraph 76 covers customer list uploads to Snap. Paragraph 77 is the roster, and Pinterest is on it. Paragraph 78 goes to knowledge, noting that Hims' SEC filings acknowledged privacy and consumer-protection regulatory risk starting in 2021 and that the FTC served a Civil Investigative Demand in October 2023.

How Pinterest Conversion Tracking Actually Works

Pinterest's measurement stack has three layers, and healthcare advertisers tend to have all three on without having decided to.

The Pinterest Tag. A base snippet loads on every page and fires a page visit. Event snippets fire on specific actions using predefined names such as PageVisit, ViewCategory, Search, Checkout, Signup and Lead, plus custom events, and each can carry parameters. The base snippet alone already sends the page URL and referrer, which on a telehealth site is frequently the whole story. A URL like /weight-loss/consultation/confirmation does not need a payload to be revealing.

Automatic event tracking. Pinterest offers an enhanced setup that detects and fires events based on page behavior without an engineer mapping each one. This is a convenience feature and a compliance hazard at the same time. It means the set of events being transmitted can change without a code deployment, without a ticket, and without anyone in your organization knowing that it changed. Any tracking configuration that can expand itself belongs in a monthly review, not an annual one.

Enhanced match and the Conversions API. Enhanced match passes a hashed identifier, typically an email address, so Pinterest can tie the site action to a Pinterest account. The Conversions API sends the same conceptual payload from your server rather than the visitor's browser, usually with an event ID for deduplication against the browser tag. Hashing is not de-identification here. A SHA-256 email is a stable, portable join key, which is what makes it valuable to the platform and what makes it a disclosure. Our walkthrough of auditing pixels for PHI leakage covers how to trace these fields in a live browser session.

How Curve Handles Pinterest Tracking

Curve sits between your site and Pinterest rather than letting Pinterest's code sit on your pages. Events are captured first-party, then sanitized on Curve's servers before anything is sent onward, so identifiers, condition-bearing URL paths and free-text fields are stripped or replaced rather than hashed and forwarded. What reaches Pinterest is a conversion signal with the campaign and click context the platform needs for optimization, and without the protected health information it does not need and should never hold. Destinations are configured per platform, so a Pinterest setup can carry a different, narrower payload than a search destination that legitimately needs more. Curve signs a business associate agreement, which is the baseline no ad platform will offer you, and which is why the sanitization has to happen before the handoff rather than after. If you are building the Pinterest side of a broader program, our guide to Pinterest marketing for wellness practices covers the campaign structure that pairs with this.

Why Discovery Traffic Is Unusually Revealing

Most pixel analysis treats a page view as weak signal and a purchase as strong signal. On a discovery platform that intuition inverts. Search intent is transactional and often shallow, and social retargeting intent is borrowed, since the person is only on the site because you followed them there. Pinterest intent is neither. Pinterest is used as a planning surface, where people assemble boards months before they act, and the categories that dominate the platform are the same ones that dominate telehealth: weight management, skin, hair, fertility and pregnancy, menopause, sleep, anxiety and sobriety.

The practical consequence is that a Pinterest referral landing on a condition-specific page carries more inference than the same landing from a broad search campaign. The user did not stumble in. They were saving ideas about the thing, then clicked through to a provider that treats the thing. A single PageVisit event, with a URL and a hashed email attached, is enough to place that person in a condition category with high confidence.

This is the logic behind paragraph 74. The complaint is not upset that audiences existed. It alleges the audiences were specific, that the specificity had to come from somewhere, and that where it came from was user health information the company had promised to keep private. Discovery traffic is where that specificity is cheapest to acquire.

Who Is Running Pinterest Ads in Healthcare Right Now

Pinterest campaigns in this space cluster into a few recognizable groups, and each has a different exposure profile.

• Weight management and GLP-1 telehealth. The heaviest current spenders, and the most scrutinized category on every platform, where the ad policy layer and the privacy layer are both moving at once.
• Aesthetics, dermatology and med spas. Visual before-and-after content performs, which pulls these advertisers onto Pinterest early and often through an agency that owns the tag.
• Fertility, IVF and maternal health. The most sensitive category on the platform by a wide margin, and the one where a single conversion event maps to a diagnosis with almost no ambiguity.
• Hair, sexual health and mental wellness. The stigma-adjacent categories, where the privacy promise is doing the most marketing work and therefore where a broken promise is most legally exposed.

In most of these organizations, nobody who understands HIPAA has ever looked at the Pinterest tag. An agency pasted it into a tag manager container during a campaign launch and it has been firing ever since.

What to Check on Your Own Pinterest Setup

Six things, in the order that finds problems fastest.

1. Inventory where the tag fires. Open a browser network panel, filter for Pinterest's collection endpoint, and walk your funnel from home page to condition page to intake to confirmation. Most teams are surprised at the confirmation page.
2. Read the URLs you are sending. The page path and referrer travel by default. If your information architecture encodes the condition in the path, and most telehealth architectures do, you are transmitting the diagnosis in the plainest possible form.
3. Check whether enhanced match is on. It is frequently enabled at the account level. Find out whether a hashed email is going out with your events, and from which pages.
4. Check whether automatic event tracking is on. If Pinterest is inferring events for you, your transmitted event set is not the one in your documentation.
5. Look at your audience definitions. An audience named after a condition documents an inference about the people in it. That is the paragraph 74 exposure in its clearest form, and there are ways to build reach without it, covered in our piece on building lookalike audiences from in-market segments without PHI.
6. Find out who owns the tag. If the answer is an agency, or nobody, fix that first. Every other control depends on someone being accountable for what the container ships.

The Server-Side Question, Answered Directly

Advertisers routinely ask whether moving Pinterest to the Conversions API resolves the exposure. It does not, and the complaint is the reason we can say that with confidence rather than as a prediction. Paragraph 70 shows the FTC describing Meta's server-side connection correctly and treating it as a sharing vector regardless, and paragraph 77 names Google Ads S2S and TikTok s2s explicitly. The agency is not distinguishing between transport mechanisms. It is looking at what information ended up at an ad platform, and whether the company had promised its users that would not happen.

Server-side transport is genuinely valuable, because it is the only place you can inspect and modify a payload before it leaves. That control is the point. The transport by itself does nothing. A server-side integration that forwards the same URLs and the same hashed emails is the browser pixel with extra steps and a better latency profile. The distinction is developed further in our analysis of whether the Meta Pixel or Conversions API is HIPAA safe, and the reasoning transfers to Pinterest without modification.

Frequently Asked Questions

Does the FTC complaint say Pinterest received health data from Hims?

The complaint alleges that Pinterest tracking code was among the third-party pixels placed on Hims platforms (paragraph 77), and that pixels on those platforms transmitted Events, defined as visitor actions on the website, to advertising platforms. The complaint does not single Pinterest out for separate treatment. These are allegations that have not been tested in court, and Hims has denied them and says it will defend the case.

Is the Pinterest tag HIPAA compliant?

No ad platform tag is HIPAA compliant on its own, and Pinterest does not offer a business associate agreement to advertisers. Compliance is a property of your configuration, not of the vendor. If your setup can send condition-bearing URLs or hashed identifiers from pages tied to a treatment category, the tag is a disclosure mechanism regardless of what the platform's own policies say about health data.

Can healthcare companies advertise on Pinterest at all?

Yes. Pinterest has ad policies restricting certain health claims and products, but the platform is open to healthcare advertisers, and the discovery format works well for the categories that dominate telehealth. The issue is never the ad. It is what the conversion tracking sends back.

Does using Pinterest's Conversions API instead of the browser tag fix the problem?

Not by itself. The FTC described Meta's server-side Conversions API accurately in paragraph 70 and pled it as a sharing vector anyway, and paragraph 77 names two other server-side integrations. Server-side transport is useful because it gives you a place to sanitize the payload, but if you forward the same data, you have changed the plumbing and nothing else.

How does this compare to GoodRx and BetterHelp?

Those cases settled. GoodRx settled with the FTC in February 2023 for $1.5 million, BetterHelp in March 2023 for what ultimately reached $7.8 million. The Hims & Hers matter is being litigated rather than settled, seeks civil penalties, and has California and Utah as co-plaintiffs. Our mid-2026 roundup of pixel settlements tracks the full progression.

This article reflects the public record as of July 2026, and describes allegations in a complaint that has not been adjudicated. The full filing is available as the redacted e-filed complaint on ftc.gov.

If you are running Pinterest campaigns for a healthcare or telehealth brand and cannot currently say, from documentation rather than memory, exactly what your tag transmits from your confirmation pages, that is the gap worth closing this quarter. Curve gives healthcare advertisers conversion tracking that keeps Pinterest optimizing on real signal while protected health information stays on your side of the line, with a business associate agreement in place. See how it works at curvecompliance.com.