Judge Barker (E.D. Tex.) Disassembles Two SAD Scheme Cases

October 5, 2026 · by · in E-Commerce, Patents

Judge J. Campbell Barker (E.D. Tex.) issued two rulings involving SAD Scheme patent cases last week. The cases are:

  • Shangyou Jiayi Lighting Product Co., Ltd. v. Schedule A Defendants, 2026 WL 2948520 (E.D. Tex. Sept. 30, 2026) and
  • Haiying Liu  v. Schedule A Defendants, 2026 WL 2948569 (E.D. Tex. Sept. 30, 2026)

As per the new norm, these cases appear to be Chinese companies suing their Chinese competitors in US court. Another example of how rapidly the Chinese learn from U.S. (judicial) “innovations.”

Judge Barker issued two very similar opinions. Neither went well for the plaintiffs. Many of his points fundamentally reject standard SAD Scheme arguments that are prevalent in many other cases.

Given their extensive overlap, I’ll cover the Shangyou Jiayi Lighting case first and then discuss some differences in the Liu opinion.

Shangyou Jiayi Lighting

Judge Barker starts with an obvious yet often undervalued point: “Schedule A cases can have an uncertain fit with the Federal Rules of Civil Procedure and due process.” I would have gone further and said the cases categorically don’t comply with FRCP or due process, but at least this shows he understands the gaming risk.

Email Service of Process

“Two circuits have now held that the Hague Convention, when it applies, prohibits service by email on defendants in China. The court agrees with that analysis….The court agrees with the Second Circuit’s recent conclusion that, where the Hague Convention governs service of process on a defendant in China, it prohibits service by email. ”

The plaintiff weakly argued that it was hard to find the defendants’ physical addresses. That did not go well:

Plaintiff argues that defendants primarily use electronic communications to operate their businesses and communicate with customers and third-party sales platforms. That answers the wrong question. Conducting business online and communicating mainly by email bear on whether email is reasonably calculated to give notice under Rule 4(f)(2) or (3) and the Due Process Clause…It says nothing about whether any defendant’s physical address is known. And only that question controls whether the Convention applies here.

A defendant may do business entirely by email and still have a physical address that is known or discoverable with diligence….

Diligent efforts to find a defendant’s address include more than just stating the difficulty of the endeavor….

Plaintiff has put forth no evidence that it could not find defendants’ Chinese addresses despite diligence. So plaintiff has not established this exception to the Hague Convention’s application.

Personal Jurisdiction

“Even the approaches most favorable to Schedule A plaintiffs rely on a completed sale of an accused product into the forum. As explained below, the record here shows no such sales by any defendants.”

The plaintiff’s items of evidence “are screenshots of Amazon checkout pages listing an address in Plano, Texas, which is within this district, but no completed purchases. Pages showing that a product could be shipped to the forum do not establish a sale into it.”

Asset Freeze

As usual, the plaintiff sought to freeze of the defendants’ money held in third-party accounts and obligate third-party services to freeze the money. The plaintiff backed off the freeze order on the third-party services, but the court also highlights a jurisdictional problem: “To the extent the proposed order would have been binding on the platforms directly, apart from any relationship to defendants, it would still have adjudicated defendants’ interests in their own property without jurisdiction over them.”

Unmasking Subpoenas

Judge Barker rejects a general discovery order to third-party services. However, “If plaintiff believes that a platform’s records would show sales of accused products shipped to Texas, it may move for leave to serve targeted subpoenas for that limited purpose.”

Joinder

“Schedule A cases often join dozens or hundreds of unrelated online sellers in a single action based on little more than their alleged infringement of the same intellectual property. Courts have increasingly rejected that practice.”

Here, the plaintiff narrowed the defendant list to the point where the judge accepts joinder:

The concern that can arise with Schedule A joinder in other cases is not present here because plaintiff does not sweep together unrelated sellers. It sues sales channels allegedly operating together under one operation. Plaintiff supports the connection among the defendants with allegations beyond the alleged infringement itself: use of the same or similar advertising look and feel, product images and descriptions, and tactics to evade enforcement efforts. On plaintiff’s theory, its claims against each defendant arise from the sale of the same accused products by the same operation and will present common questions of fact. That satisfies § 299(a) at this stage.

I don’t know about this. SAD Scheme plaintiffs routinely claim that the defendants all use the “same or similar advertising look and feel, product images and descriptions, and tactics to evade enforcement efforts.” To me, this looks like a standard plaintiff boilerplate roboargument.

Sealing of Filings

Temporary sealing can serve a legitimate purpose in an ex parte proceeding. It can prevent a defendant from defeating relief before the court can act. But that justification lasts only as long as the application for ex parte relief is pending, and it ends once the court has ruled. The court has now denied the temporary restraining order and asset restraint. Now that the court has ruled, the public’s interest in access extends to the ruling itself and to the record on which it rests. And no document sought to be sealed is claimed to contain confidential business information or personal information warranting protection.

Continued sealing would also undermine protections for defendants in the federal rules. A plaintiff may voluntarily dismiss an action without a court order before defendants answer. But the federal rules protect defendants against repeated litigation of the same claim: A second voluntary dismissal of the same claim operates as an adjudication on the merits. And a court in which the claim is refiled may award the costs of the prior action and stay proceedings until they are paid.

Those protections, and the related-case disclosures that courts require when a case is filed, depend on the prior action being discoverable. If this record remained sealed and the case were dismissed by plaintiff, neither defendants nor another court could readily learn that the claims had been asserted or that relief had been sought and denied. That concern is not hypothetical in Schedule A litigation.

It’s great that Judge Barker recognizes the risks of plaintiffs dragging defendants through multiple lawsuits until the plaintiffs get the positioning they want. Docket transparency is a partial cure, but more is needed too.

Notice to Defendants of the TRO Request

The plaintiff didn’t notify the defendants of the TRO request and didn’t adequately explain why not:

The reasons offered for proceeding without notice are generic. Plaintiff says only that defendants “can and likely will take immediate steps to permanently alter the status quo, including but not limited to steps such as registering new e-commerce stores under new aliases and moving any assets to offshore bank accounts outside the jurisdiction of this Court.” Nothing in the record describes anything these defendants have done to suggest they would dissipate assets or evade an order.

Judge Barker blows up the standard SAD Scheme plaintiff expectation that it’s obvious every Chinese defendant will go rogue. That claim requires more facts than just rote recitation.

Likelihood of Success on the Merits

Judge Barker notes the problems with testing the claim’s legitimacy in an ex parte proceeding:

Ordinarily, an accused infringer can defeat preliminary relief by raising a “substantial question” of infringement or validity. But, here, no defendant is present to do so. That makes it all the more important that the court at least attempt to test plaintiff’s showing on its own terms—and make its assessment of likelihood of success with eyes wide open to the limitations of a one-sided presentation.

This is true with utility patent claims and even more true for design patent claims. Our adversarial system depends on defendants calling attention to the weak points of the plaintiffs’ claims. Judges are not well-positioned to intuit what defendants might hypothetically say if they were around.

With respect to claim 1 of the patent:

The claim chart features only a single unidentified product, which plaintiff alleges—with no declaration, affidavit, or legal analysis—is “representative of all of the Infringing Products identified in Schedule A”. The photographs of the unidentified product contain a few labels purporting to identify only a lamp, with its conducting-wires, light-emitting components, conductor and insulating layers, plurality of openings, and conduct-or-contact surfaces. This is facially insufficient to establish a likelihood of success on the merits where the limitations of Claim 1 are far more substantial than the limited number of identified details in each photo. In fact, plaintiff’s showing is scarcely more than the screenshot-only presentations that have drawn criticism in other Schedule A cases….

the court cannot comprehend how a chart exemplifying a mere “perfunctory submission” can be said to show a likelihood of success on the merits.

The chart refers to “Defendant #10 Infringing Product” in the singular, and it does not say which listing the photographed product came from or how plaintiff examined it. But plaintiff accuses 14 separate products of infringement. Nothing in the record shows that each accused product contains the structure depicted in the chart, except for plaintiff’s assertion that they are “identical … in all aspects relevant to the [patent] claims.”…

“[S]imply mimicking the language of the claims when identifying infringement” fails to establish even reasonable notice of plaintiff’s theories of infringement, and it is certainly insufficient to find a likelihood of success on the merits….

The product identifiers listed in the motion cover some 14 listings. Plaintiff has not shown a likelihood of infringement as to any of them.

Asset Freeze Redux

“because a utility patent plaintiff’s monetary remedy is legal damages, a federal court lacks authority to freeze a defendant’s assets to preserve that remedy.” [cite to Grupo Mexicano de Desarrollo]

Judge Barker also explains how an asset freeze would reach assets unrelated to the alleged infringement:

an asset freeze preserves only assets that may be needed to satisfy that equitable relief. But the proposed order would bar defendants from transferring “any money held by a Third Party Provider,” without regard to location, amount, or source. Plaintiff now concedes that defendants may continue selling non-accused products. Yet the proposed order on file would reach the proceeds of those very sales. And it would freeze all funds associated with defendants’ storefronts and with any other accounts later identified….

The requested freeze would thus restrain defendants’ entire business to secure a claim for damages that plaintiff has not attempted to quantify.

Overfreezes are ubiquitous with the SAD Scheme. The plaintiffs usually cannot model how much money is atttributable the defendant’s infringing conduct (if any) at the ex parte TRO stage (pre-discovery). Thus, the TRO routinely freeze all of the defendant’s assets, including proceeds of non-infringing conduct. That is categorically a due process violation.

Echoing the Eicher Motors case, Judge Barker flags the harms that an asset freeze can cause (emphasis added):

The requested freeze would do more than preserve the status quo. Defendants would first learn of this suit when their accounts were frozen. The freeze would reach “any money held by a Third Party Provider,” including the proceeds of products that plaintiff concedes defendants may keep selling. Such a freeze “locks down defendants’ assets” and can cause “severe or fatal cashflow problems for the defendant, which may not be able to pay its vendors, employees, or lawyers.”…

Without defendants’ participation, moreover, the court has no reliable way to weigh the harm that the freeze would cause them or plaintiff’s likelihood of success in overcoming any invalidity or non infringement defense. An asset restraint that impairs a defendant’s ability to answer the claims against it is not a measure that preserves the status quo pending a hearing. It is a measure that may prevent a hearing from ever occurring.

This is such an obvious point, and yet judges often elide it. Ex parte TROs are supposed to temporarily preserve the status quo, but SAD Scheme ex parte TROs routinely disrupt the status quo and change the parties’ substantive rights. That is another part of their due process problems.

Bond

The court doesn’t grant the equitable relief, so no bond was required. Nevertheless, Judge Barker pushes back on the plaintiff’s request:

The proposed order would freeze “any money held by a Third Party Provider” of 14 defendants. It would bar sales of some 14 product listings. The court is not satisfied that a mere $10,000 injunction bond would properly compensate defendants in the event that the requested TRO, if entered, turns out to have been legally or factually improper.

In my SAD Scheme Standing Orders paper, I note that judges don’t have any rigorous methodology for setting bond amounts. This is especially a problem in ex parte proceedings, where the judge has no indication of how the ex parte TRO will impact defendants.

Implications

Reading this opinion, this meme kept coming to mind:

In my ideal world, Judge Barker would have rejected the entire premise of the SAD Scheme cases, rather than poking holes in multiple individual pieces. Even so, this opinion calls out many problems with standard SAD Scheme lawsuits. Many of those problems are essentially manifestations of an overdeveloped sense of entitlement, i.e., “I’m an IP owner, so I deserve everything I ask for without question.” Fortunately, Judge Barker doesn’t rubberstamp those requests.

Liu Case

The Liu decision has a lot of overlapping language with the Shangyou Jiayi Lighting decision, so I will just highlight some of the new incremental points of interest.

Personal Jurisdiction

The only evidence the motion offers for those assertions is one sentence in plaintiff’s declaration, stating that defendants “collectively advertise, distribute, and sell the accused product into the United States, including Texas.”

That sentence identifies no sale, buyer, shipment, or date in Texas. It attributes the conduct to defendants collectively rather than to any one of them. And it has no stated basis. Plaintiff resides in Shaanxi Province, and the declaration refers only to “my investigation,” without saying what that investigation involved….

Exhibit 1 is a Facebook page for “Tiny Land Inc” that lists Shanghai Tailan as the entity responsible for the page. Exhibit 2 consists of photographs of a product box. Exhibit 3 is a Chinese corporate-records printout for Shanghai Tailan. None of these materials mentions a sale, a shipment, or a customer in Texas….

For the tinylandus.com website and the Wayfair store-front, plaintiff’s exhibits include only checkout pages listing Sherman, Texas, shipping addresses, with no completed purchase. For the Walmart storefront, the exhibits show only items placed in a shopping cart. Pages showing that a product could be shipped to the forum do not establish a sale into it.

Though the plaintiff didn’t highlight it, there was one sale delivered to a Texas address. The court says that evidence might provide enough support for jurisdiction for that sale.

Joinder. “Plaintiff supports the connection among the defendants with evidence beyond the alleged infringement itself: the Facebook page, the product packaging, and the shared email domain.” I’ve seen weaker evidence in support of joinder, but this could also be a roboargument.

Notice before the ex parte request. As further evidence that the plaintiff didn’t have any reason to assume asset dissipation on notice, the court says “The business operator that plaintiff identifies is, per plaintiff’s own evidence, a registered company with a known address that has operated under the same name for years.”

Likelihood of success on the merits.

With respect to claims 1 and 8:

Plaintiff argues that the accused products “practice each and every limitation of at least independent claims 1, 8 and 9” of the asserted patent. But plaintiff’s claim chart addresses only claim 9. The chart expressly states that plaintiff “reserves the right” to assert claims 1 and 8 later, “based on further investigation and discovery.”

I’m pretty sure that’s not how pleading burdens in patent cases are supposed to work, but in SAD Scheme cases, plaintiffs often get away with a lot.

With respect to claim 9:

Plaintiff’s chart maps each limitation of claim 9 to photographs of a physical product. The photographs include labeled images of what plaintiff identifies as the annular and protective sleeving pipes, and images with a measuring tape comparing framering diameters before and after folding. That is a more substantial showing than the screenshot-only presentations that have drawn criticism in other Schedule A cases.

It’s amazing how much judges appreciate the plaintiffs taking small but seemingly obvious information steps like labeling photos, given how often SAD Scheme plaintiffs cut procedural corners and impose significant workloads on the judge’s staff. It’s still not enough in this case:

First, it concerns a single, unidentified product. The chart refers to “Defendants’ Accused Product” in the singular, and neither the chart nor plaintiff’s declaration says which listing the photographed product came from, who obtained or examined it, or how. Plaintiff accuses 20 listings across four sales channels, identified by seven Amazon ASINs, three Walmart product IDs, three Wayfair product IDs, and seven website product IDs. These include a standalone 69-inch crawl tunnel and multipiece sets combining tents, ball pits, and tunnels of different sizes. Nothing in the record shows that each accused product contains the structure depicted in the chart.

Second, several of the chart’s key entries restate the claim language without explaining how the photographed structure meets it….

Third, plaintiff’s validity showing consists entirely of the statutory presumption. That presumption does carry weight at this stage. But on an ex parte record, the court has no way to assess prior art or other alidity challenges a defendant might raise.

This looks like another IP Privilege card. “I have a patent, so I win.”

Judge Barker summarizes the discussion:

At most, therefore, plaintiff has made a preliminary showing that the single product photographed in its claim chart likely meets the limitations of claim 9 and no patent-specific showing as to its likely validity. It has made no showing as to claims 1 and 8 and no showing as to any other accused product.

Bonus: The Smiley Company SPRL v. Schedule A Defendants, 1:25-cv-26045-RS (S.D. Fla. Sept. 23, 2026).

A Smiley case is likely going to fall apart due to misjoinder:

counsel for Plaintiff informed the Court that Defendants are likely located in the same region, China, and are operating in the same manner by selling infringing products online. Counsel for Plaintiff did not set forth any specific allegations explaining how the 50 remaining Defendants were connected to one another.

Simply alleging that Defendants have violated Plaintiff’s trademark in the same way is insufficient.

Great. At the same time, 32 defendants in that case have already been dismissed, likely because they settled a case that apparently violated joinder rules from day 1. Will the court take a look at the (il)legitimacy of those dismissals and any resulting settlements? Or did the plaintiff already reap the economic spoils from those 32 defendants based on an ex parte TRO that never should have issued?

Prior Blog Posts on the SAD Scheme