ConsentPixel – Privacy · Verified

Home › Meta Pixel Lawsuit Tracker › Baker v. Seattle Children's Hospital
Case Deep-Dive · Meta Pixel / Healthcare · State Supreme Court

Baker v. Seattle Children's Hospital: a Meta Pixel win that's narrower than it looks

On October 8, 2026, the Washington Supreme Court handed a hospital a clean victory in a Meta Pixel privacy class action — holding that the state's all-party-consent wiretap law simply doesn't reach what the pixel did on Seattle Children's public website. For defendants, it's a real and useful win, and the headlines read "hospital not liable."

But read the opinion and a different story emerges. The court won on a narrow point of statutory grammar — not on whether tracking is private, not on whether visitors consented, and explicitly not on patient portals, live chat, or any channel where a human is on the other end. A justice writing separately would have held the opposite on the very question the majority sidestepped. This is a case about which word a statute uses — and about how much of the risk it leaves untouched.

✍️ By the ConsentPixel Team 📅 Published November 2026 ⏱️ ~14 min read ⚖️ Status: dismissal affirmed (Oct 8, 2026)
RCW 9.73.030
The Washington Privacy Act provision the whole case turned on — all-party-consent wiretapping
1 word
"individuals" — the statutory term that decided the outcome
Public site only
Plaintiffs dropped their MyChart patient-portal claims — the authenticated-page question is still open

⚖️ Case snapshot

Case
Baker v. Seattle Children's Hospital
Court
Washington Supreme Court (en banc)
Case number
No. 104590-5
Author
Angelis, J. (majority); González, J. (concurring in result)
Decided
October 8, 2026
Defendant
Seattle Children's Hospital (Washington nonprofit corporation)
Plaintiffs
Carly Baker, Janssen Ramos Savoie, Amber Shavies — parents of patients, each with a Meta account
Legal theory
Washington Privacy Act, RCW 9.73.030(1)(a) (the only claim on appeal)
Tracking tech
Meta Pixel on the hospital's public website; _fbp and c_user cookies
Settlement
None — defense win on the merits of the statute
Status
Dismissal affirmed. WPA claim fails; court did not reach privacy, consent, portals, or the other (unappealed) claims
Significance
A narrow statutory-interpretation win that leaves most pixel-litigation theories intact
Not legal advice. This article is an educational summary of the published opinion in Baker v. Seattle Children's Hospital, No. 104590-5 (Wash. Oct. 8, 2026), and legal reporting about it. Allegations described are claims made in litigation, not findings of fact; the decision interprets one Washington statute and does not resolve every theory or apply outside Washington. It is not legal advice — consult a qualified attorney about your own situation.

What the case is about

Three parents — Carly Baker, Janssen Ramos Savoie, and Amber Shavies — each had a child who was a patient at Seattle Children's Hospital, and each had a Facebook or Instagram account. In a putative class action, they alleged that the hospital's public website carried the Meta Pixel, Facebook's tracking code, which copied what they did on the site and sent it to Meta without meaningful disclosure.

The specifics are what give the case its emotional weight. According to the complaint, Baker searched the site for medical conditions and symptoms; Ramos Savoie searched and used the "Find A Doctor" page; and Shavies looked up urgent-care locations and their hours. The pixel allegedly relayed those page requests to Meta — the complaint's vivid example is that typing "depression" into the site's search bar would be transmitted to Meta and linked to the visitor's account for ad targeting. Some parents said they later saw health-related ads on Facebook.

Two cookies sat at the center of the mechanics. An _fbp cookie grouped a visitor's activity under a device identifier, and a c_user cookie tied that activity to a named Facebook account. Visitors without a Meta account were allegedly still tracked through a unique identifier that could be connected to them later if they ever logged in. The hospital's cookie pop-up told visitors that accepting cookies meant agreeing to software that stores data to assist marketing — but it did not name the Meta Pixel or explain what it did.

It is, in other words, the archetypal healthcare pixel complaint: a parent researches a child's medical condition on a trusted hospital's website, a tracker fires, and the sensitive query allegedly lands in an advertising giant's profile of that person. What makes Baker matter isn't the fact pattern — it's the specific, and unusually technical, reason the claim failed.

The legal theory — Washington's all-party-consent law

The plaintiffs did not sue under California's CIPA. They sued under the Washington Privacy Act (WPA), specifically RCW 9.73.030(1)(a). Washington is one of roughly eleven "all-party consent" (or "two-party consent") states, where intercepting or recording a protected communication requires the consent of every participant — a stricter rule than the federal one-party-consent standard. On its face, that makes Washington an attractive venue for a tracking plaintiff: if a pixel "intercepts" a communication, everyone in that communication has to have agreed.

The statute protects a "[p]rivate communication . . . between two or more individuals." That phrase contains two separate requirements that a plaintiff has to satisfy: the communication must be (1) private, and (2) between two or more individuals. The plaintiffs argued their searches were private communications intercepted by the pixel without their consent. The hospital argued they failed both prongs.

Here is the move that decided the case: the court chose to resolve it entirely on the second requirement — "between two or more individuals" — and never reached the first. That choice is the whole story, because the second requirement turned out to be a grammatical dead end for the plaintiffs, while the first (whether a health search is "private") is the live, unresolved question that every other pixel case is really about.

Why "all-party consent" states matter in pixel litigation. Most website-tracking wiretap theories need a state whose law is strict enough that routing data to a third party without consent looks like illegal interception. Washington, California, Florida, Pennsylvania, Illinois and a handful of others qualify. Baker didn't weaken that strategy in general — it closed one specific door in one state, on one statutory word, for one kind of interaction (automated, machine-on-the-other-end). The strategy itself is intact.

The holding — a win built on one word

Justice Angelis, writing for the majority, affirmed the dismissal with a holding you could fit on an index card: "the communications at issue do not constitute a communication 'between two or more individuals' under RCW 9.73.030(1)(a)." The reasoning is a clinic in statutory interpretation, and it runs like this:

  1. The statute lists "individual" separately from other entities. Subsection (1) of the WPA refers to an "individual" and, elsewhere, to a "partnership, corporation, association," and other organizations. When a legislature uses different words in the same statute, courts presume it meant different things by them.
  2. So "individual" does not include a corporation. Because the legislature distinguished individuals from corporations, the court read "individual" in subsection (1)(a) to mean a natural person — not a company.
  3. Seattle Children's is a corporation, and the visitor was talking to its machine. Each alleged communication involved one plaintiff on one side and an automated system that the hospital (a corporation) owns and operates on the other. That is not a communication "between two or more individuals." It is, at most, a communication between one individual and a corporate server.
  4. Therefore the WPA does not reach it. With only one "individual" party to each communication, the statutory element fails as a matter of law, and the claim was properly dismissed.

Notice what the court did not have to decide to get there. It didn't decide whether a search for "depression" is a "private communication." It didn't decide whether the cookie banner was adequate consent. It didn't decide whether the data was sensitive, or whether Meta's receipt of it was an "interception." It resolved the case purely on who (or what) was on the other end of the line — and a corporation's automated website, the majority held, isn't an "individual."

The court also pointedly declined to import California or federal wiretap case law, citing Washington's own precedent (Roden) on the differences between the statutory schemes, and reading Roden's description of Washington's law as "the most restrictive" as being about the all-party-consent rule specifically (per Kipp), not a license to borrow other states' pixel analyses. That matters: it means Baker is a Washington-specific interpretation, not a template other states must follow.

The concurrence that proves the point

If you want evidence that this win is narrow, you don't have to look past the same opinion. Justice González concurred only in the result — he agreed the case should be dismissed, but he disagreed with almost everything that got the majority there, and he would have decided the hard questions the other way.

  • He would hold the searches are private communications. On this record, González wrote, the clicks and searches carried a reasonable expectation of privacy — exactly the prong the majority refused to reach.
  • He would protect communications made through a corporation's people. Because a corporation can only ever act through individuals, González would hold that a visitor communicating with staff acting for the corporation is protected — rejecting the majority's "individual ≠ corporation" grammar.
  • He concurred anyway — on a narrow ground. His reason for still affirming: these particular searches were directed at an automated server, not at another person. So on these facts, dismissal was right — but the door he'd leave open is enormous.

Read González's concurrence next to the majority and the lesson is unmistakable: change one fact — put a human on the other end of the interaction, as in a live chat, a patient-portal message, or a staffed contact form — and at least one justice on this court thinks the WPA would apply. The majority expressly reserved that question (in footnote 6). The win was real; the margin was one factual step wide.

Is a pixel firing on a page where someone talks to you? Baker turned on the difference between a visitor talking to a machine and a visitor talking to a person. Live chat, intake forms, portal messages — those are the pages where this win doesn't reach. ConsentPixel's free scanner loads your site as a first-time visitor and shows you what fires on every page, before anyone consents.

Scan your site free →

Five things this ruling did not decide

Because the coverage framed Baker as "hospital wins, pixel is fine," it's worth being precise about everything the decision left open. Each of these is a live avenue a plaintiff can still travel.

The court did NOT decide…Why it still matters
Whether a health search is a "private communication"The majority skipped the privacy prong entirely; the concurrence would have found privacy. It's unresolved — and it's the core question in most pixel cases.
Whether the cookie banner was valid consentThe majority noted the pop-up didn't name the Pixel and called the notice insufficient, but did not rule on whether a banner defeats a privacy expectation. Consent remains contestable.
Authenticated / patient-portal pages (MyChart)Plaintiffs voluntarily dropped their MyChart claims, so the court reviewed only the public site. Logged-in portal tracking was not blessed.
Human-to-human channels (chat, messaging, staff)Footnote 6 reserved whether the WPA applies when a corporation communicates through individuals. The concurrence says it would.
The plaintiffs' other claimsConsumer Protection Act, Uniform Health Care Information Act, invasion-of-privacy and common-law claims were not appealed, so the court never addressed them. They were not rejected on the merits.

Put those together and the honest summary is: Seattle Children's won dismissal of one statutory theory, about one kind of page (public, automated), in one state, while the court left the privacy question, the consent question, the portal question, the human-channel question, and four other claims entirely untouched. That's a defendant's win worth having — and a long way from "tracking is legal now."

Why the portal line is the one to watch

The single most important detail in Baker is one the plaintiffs handed the hospital for free: they disclaimed their MyChart patient-portal claims and limited the case to the public website. That tactical choice is why the opinion reads the way it does — and it quietly marks the exact boundary of the win.

Think about the difference. On a public hospital website, a visitor browsing and searching is interacting with an automated system — the fact pattern the majority said falls outside the WPA. But a patient portal like MyChart is a different animal: a logged-in, authenticated environment where a patient often messages a care team, sends a question to a nurse, requests a refill, or reviews results. Those are communications that can plausibly be to a person — the human-on-the-other-end scenario the majority expressly reserved and the concurrence would protect.

This tracks the national regulatory line almost exactly. HHS's Office for Civil Rights has drawn its sharpest distinction between unauthenticated public pages and authenticated pages behind a login, treating the latter as presumptively tied to protected health information. A 2024 federal ruling (AHA v. Becerra) narrowed part of OCR's guidance for public pages, but the authenticated-portal exposure survived. Baker comes out the same way by a completely different route: the public site is where the defense won, and the portal is where everyone — OCR, the concurrence, and the plaintiffs' own litigation strategy — agrees the real risk lives.

The practical reading. If your highest-risk pages are public, informational, and automated, Baker is genuinely reassuring in Washington. If your highest-risk pages are logged-in — a portal, an account area, a messaging center, a scheduling tool tied to an identity — Baker says almost nothing in your favor, and the concurrence actively cuts against you. Knowing which of your pages are which is the entire exercise.

How Baker fits the 2026 landscape

2026 has produced a run of defense-friendly headlines — SB 690 narrowing CIPA in California, the Evans/FindLaw dismissal, and now Baker. Taken together they can read like a tide turning against tracking plaintiffs. But each win is narrower than its headline, and the broader current is still running the other way. Set Baker against the rest of the year:

DevelopmentDirectionWhat it actually changed
Baker v. Seattle Children's (WA Sup. Ct., Oct 2026)DefenseWPA doesn't reach automated public-site interactions — one statute, one state, public pages only
Rand v. Eyemart Express (5th Cir., Oct 2026)PlaintiffRevived a Meta Pixel health-data suit on standing — a published federal appellate win for plaintiffs
SB 690 (CA, eff. Jan 1 2027)DefenseRemoves the private right of action for CIPA §638.51 pen-register website claims only
Frasco v. Flo Health (N.D. Cal.)Plaintiff$59.5M settlement + an August 2025 CIPA jury verdict against Meta over health-app data
Salazar v. Paramount (U.S. Sup. Ct.)WatchThe VPPA at the Supreme Court — the Court appeared likely to let the pixel suit proceed

The pattern is not "tracking litigation is dying." It's "the menu is being re-sorted." Cheap, easy claims (the CIPA pen-register theory) are getting harder; fact-specific, statute-specific claims are being refined; and appellate courts are, on balance, keeping the serious ones alive. Baker is a defense win that tells plaintiffs exactly where not to file in Washington — and, by implication, exactly where to file instead.

What this means for your site

Whether you're a hospital, a clinic, or any business whose site touches sensitive topics, the durable lesson of Baker isn't "pixels are safe." It's that your exposure depends on which pages run the tracker and who's on the other end of the interaction. Four habits follow directly from the case:

  1. Map your pages by "machine vs. human." Baker's win covers automated, public, informational pages. Your real risk concentrates where a visitor communicates with a person — live chat, contact and intake forms, appointment messaging, and especially logged-in portals. Inventory those first; they're the pages this ruling doesn't protect.
  2. Treat authenticated pages as a separate risk tier. The plaintiffs dropped MyChart for a reason, and OCR's guidance draws the same line. Any pixel, session-recorder, or ad tag on a logged-in page deserves a far higher bar — ideally, don't run non-essential trackers there at all until consent is unambiguous.
  3. Don't rely on a generic cookie banner. The court noted Seattle Children's pop-up didn't even name the Meta Pixel. A banner that lists "marketing cookies" in the abstract, while a pixel quietly ships a visitor's "depression" search to Meta, is exactly the gap plaintiffs plead. What matters is whether the tag is actually held until consent — not whether a banner is displayed next to it.
  4. Assume the other statutes are still live. Baker is one law in one state. Your visitors are everywhere, and the federal Wiretap Act, other states' all-party-consent laws, the VPPA, HIPAA/FTC exposure, and California's CIPA (for your California traffic) don't care how Washington reads the word "individual." Build for the strictest regime that reaches your audience.

This is exactly what ConsentPixel is built for. It's a CIPA-first consent platform that enforces consent before trackers load rather than just displaying a banner next to them: the Meta Pixel and similar ad/analytics tags are governed through Google Consent Mode v2, session-recording tools (Hotjar, FullStory, Microsoft Clarity and the like) are held at the DOM level until the visitor agrees, and every consent decision is recorded with tamper-evident proof — the record you'd want if a plaintiff ever asked what fired, where, and when. It can't make a lawsuit impossible; nothing can. But it lets you see and control what loads on your public pages and your logged-in ones, and show what you did.

Key takeaways

  • It's a win on grammar, not on tracking. The WPA protects communications "between two or more individuals," and the court held a visitor plus a corporation's automated website isn't two individuals.
  • The court decided almost nothing else. It didn't rule on whether health searches are private, whether the banner was valid consent, or on patient portals — and it didn't touch four other claims that weren't appealed.
  • The concurrence would have gone the other way on privacy and on communications made through a corporation's staff — a one-fact-wide margin.
  • The portal is the real boundary. Plaintiffs dropped MyChart; the authenticated-page question — where a human is on the other end — is exactly what's left open, and it's where the risk concentrates.
  • It's Washington-only. The court refused to import CIPA or federal case law, so Baker doesn't bind other states, and the federal Wiretap Act, VPPA, HIPAA/FTC, and CIPA all remain live.

Frequently asked questions

What did the Washington Supreme Court decide in Baker v. Seattle Children's Hospital?
On October 8, 2026, the court affirmed dismissal of a Meta Pixel class action under the Washington Privacy Act (RCW 9.73.030(1)(a)). It held that the statute protects a "private communication . . . between two or more individuals," and that a visitor interacting with a hospital's automated public website is not communicating between two individuals — because the hospital is a corporation and the other "party" was an automated server. On that narrow statutory ground, the WPA claim failed. The court did not decide whether the searches were private, whether consent existed, or whether patient-portal tracking is lawful.
Does this mean the Meta Pixel is legal on healthcare websites now?
No. The ruling interprets one Washington statute and applies only in Washington. It did not hold that tracking is lawful, private, or consented — it held that this particular law's "two or more individuals" requirement wasn't met on a public, automated website. The federal Wiretap Act, other states' all-party-consent laws, the Video Privacy Protection Act, California's CIPA, and HIPAA/FTC enforcement all remain live, and a federal appellate court (the Fifth Circuit, in Rand v. Eyemart) revived a very similar Meta Pixel health-data suit the same month. This is general information, not legal advice.
Why did the case turn on the word "individuals"?
The Washington Privacy Act lists "individual" separately from "partnership, corporation, association" and other entities. Under the interpretive presumption that a legislature means different things by different words, the court read "individual" to mean a natural person, not a company. Since Seattle Children's is a corporation and each alleged communication was between one plaintiff and the hospital's automated system, there was only one "individual" involved — so the communication wasn't "between two or more individuals," and the statute didn't apply. The concurrence disagreed with this grammar and would have read "individual" more broadly.
What about patient portals like MyChart — are those covered?
The court did not decide. The plaintiffs voluntarily dropped their MyChart patient-portal claims and limited the case to the public website, so the opinion only addresses automated public pages. Authenticated portals are a different and higher-risk category: they're logged-in environments where a patient often messages a care team or provider — communications that can plausibly be "to a person," which is the human-on-the-other-end scenario the majority expressly reserved and the concurrence would protect. HHS/OCR guidance draws the same public-vs-authenticated line, treating logged-in pages as presumptively tied to protected health information. Treat portal tracking as unresolved and high-risk.
What was the concurrence, and why does it matter?
Justice González concurred in the result — he agreed the case should be dismissed — but disagreed with the majority's reasoning on the key questions. He would have held that the plaintiffs' searches were private communications, and that communications made through a corporation's staff are protected (because a corporation acts only through individuals). He concurred only because these particular searches were directed at an automated server rather than a person. It matters because it shows how narrow the win is: change one fact — put a human on the other end — and at least one justice thinks the WPA would apply.
What should my organization do in light of Baker?
Map your pages by whether a visitor interacts with a machine or a person — Baker's win covers automated public pages, but your real exposure is live chat, intake forms, messaging, and logged-in portals. Treat authenticated pages as a separate, higher-risk tier and avoid running non-essential trackers there until consent is unambiguous. Don't rely on a generic cookie banner that doesn't actually hold the tag until consent — the court noted Seattle Children's banner didn't even name the Pixel. And build for the strictest law that reaches your audience, since the federal Wiretap Act, other state wiretap laws, the VPPA, CIPA and HIPAA/FTC are all unaffected. This is informational, not legal advice — consult a qualified attorney about your own exposure.

The bottom line

Baker v. Seattle Children's Hospital is a genuine defense victory, and healthcare defendants in Washington should welcome it. But it's a victory with carefully drawn edges. The court won on a point of statutory grammar — that a corporation's automated website isn't an "individual" — and in doing so it deliberately avoided the questions that actually drive pixel litigation: whether a health search is private, whether a cookie banner is consent, and whether any of this changes once a human is on the other end of the conversation.

The plaintiffs' own decision to drop their patient-portal claims, the concurrence's willingness to decide the reserved questions against the hospital, and the Fifth Circuit reviving a nearly identical suit the same month all point to the same conclusion: the pixel fight didn't end in October 2026 — it got more precise. If your sensitive pages are public and automated, this ruling is a comfort in Washington. If they're logged-in, staffed, or outside Washington, it changes very little. Either way, the only way to know which category your pages fall into is to look at what actually fires on them. Start there.

Check your sensitive pages — free. Baker comes down to what loads, on which pages, and whether a person is on the other end. ConsentPixel's scanner loads your site as a first-time visitor and shows you exactly which trackers fire before consent — on your public pages and your logged-in ones. See it in about 30 seconds, no login, then run the full platform free for 14 days (no credit card).

Scan your site free →
CP
The ConsentPixel Team

ConsentPixel — Privacy · Verified builds prevention-first consent tooling that holds non-essential trackers until visitors genuinely consent, on any platform, and records what happened. Sources & disclaimer: this analysis summarizes the published opinion and legal reporting cited below; it is educational information, not legal advice, and ConsentPixel is not a law firm. Case outcomes are specific to their facts, statutes and jurisdictions and may be appealed or distinguished. Verify against the opinion and qualified counsel.

Sources

  1. Baker v. Seattle Children's Hospital, No. 104590-5 (Wash. Oct. 8, 2026) (en banc) — published opinion (Angelis, J.; González, J., concurring in result), Washington Courts, courts.wa.gov/opinions/pdf/1045905.pdf.
  2. Orrick, "Washington Supreme Court Rejects Privacy Act Claim Over Hospital Website Tracking Pixel: Implications for Businesses Operating Public-Facing Websites" (Oct. 2026) — orrick.com.
  3. "Parents lose Meta Pixel wiretap case against Seattle Children's Hospital," PPC Land (Oct. 2026) — ppc.land.
  4. "Washington Supreme Court Rejects Privacy Class Action And Holds That Hospital's Use Of Meta Pixel Does Not Violate State Privacy Act," Lexology (Oct. 2026) — lexology.com.
  5. Helsell Fetterman, "Washington Supreme Court: Privacy Act Doesn't Reach Hospital's Meta Pixel Use" (Oct. 9, 2026) — helsell.com.
  6. Washington Court of Appeals, Div. I, unpublished opinion affirming dismissal (Aug. 18, 2025); Wash. Privacy Act, RCW 9.73.030(1)(a).
Disclaimer: ConsentPixel provides privacy tooling and information, not legal advice. Statutory-damages figures (e.g. $5,000 under Cal. Penal Code §637.2) are statutory amounts that vary by case and are not a prediction of any outcome. Baker arises under the Washington Privacy Act (RCW 9.73.030), not CIPA, and the decision does not resolve every claim or apply outside Washington.
Scroll to Top