A deal that buys peace with one company can weaken recovery against another. The Federal Circuit has made that risk harder to leave for later.

If your company has ever settled a patent case or signed a cross-license, pull the file. The agreement that ended one lawsuit may be the document the next defendant uses to erase years of damages from a case you have not yet filed. This is not a problem reserved for companies in the business of suing on patents. It reaches any company that trades access to a large portfolio.

Picture three companies. A and B cross-license their patents to stay out of court. Years later, A sues C. C's lawyers find a B product that carries no patent number and say it practices A's patent. Now, before A can recover a dollar of pre-suit damages from C, A must show either that B's product never practiced the patent or that A made reasonable efforts to get it marked. C never signed the peace agreement between A and B. C may nonetheless walk away with a share of its value.

That is the corporate danger illuminated by VDPP, LLC v. Volkswagen Group of America, Inc., the Federal Circuit's precedential August 19, 2026 decision. The panel took a familiar marking problem and gave it force at the pleading stage. VDPP had licensed the asserted patent to eleven settling defendants. Because VDPP wanted pre-suit damages, the court held, VDPP had to plead compliance with the marking statute, section 287(a), and that duty covered its licensees' products as well as its own. Pleading that VDPP was a non-practicing entity with nothing of its own to mark was not enough. The licenses were written as ordinary licenses to make, use, and sell products. One expressly excused marking. And VDPP itself maintained that the licensed products practiced the patent. On that record the panel could see no amendment that would plausibly allege reasonable efforts to get the licensees to mark. The case, on an expired patent, ended there.^[1]

Expect to see that reasoning in a motion aimed at your own licensing history. The decision is precedential, so every district court must follow it. But precedent alone is not why it will spread. The motion costs the defendant almost nothing. In VDPP, Volkswagen did little more than point to VDPP's earlier settlement agreements in its motion to dismiss. From that moment the question was what the patent owner had done about marking, and the panel faulted the proposed amendment for saying nothing on the subject. The motion also lands before discovery, when the owner may not know which licensee products are in play, let alone whether any of them ever carried a patent number. And the triggering facts are ordinary. A license to a company that sells products the owner believes practice the patent, under an agreement that says nothing about marking, is a fair description of most cross-licenses between operating companies. An owner with dozens of those agreements and no record of marking efforts looks exactly like VDPP, only bigger.^[1]

VDPP did not invalidate cross-licenses or wipe out every company's back damages. The threat is narrower than that, and more credible for it. The price of patent peace can now include weaker enforcement against outsiders, an expensive reconstruction of a licensee's product history, and a demand for facts before discovery opens. The larger the portfolio, the more places that problem can hide.

What the panel actually decided

The holding is narrower than the headline. VDPP had eleven settlement licenses, continued to insist that the licensed products infringed, and offered only conclusory marking allegations. On that record, amendment was futile. The panel left open whether a licensor can ever show reasonable efforts without a contractual marking clause. It also found an independent procedural ground for denying amendment, and it affirmed fees on a broader record of litigation conduct.^[1]

The underlying rule predates VDPP. Section 287(a) can cut off damages when patented products sold by the patentee or its licensees carry no patent notice. In Arctic Cat, decided in 2020, an operating company lost its pre-suit damages because its licensee Honda had sold unmarked products under an agreement that excused marking. Stopping those sales did not fix it. The exposure for operating companies was already real; VDPP moved it to the front of the case.^[2][3]

Ortiz & Associates Consulting, LLC v. Vizio, Inc. supplies a related warning, and an important limit. That December 2025 decision was nonprecedential, and it reviewed only a fee award; the dismissal itself was never timely appealed. The panel discussed earlier dismissals with prejudice that the district court had treated as licenses. It did not hold that every walk-away dismissal creates an ongoing license to manufacture.^[4]

The real shift is timing. Under VDPP, the marking fight starts at the complaint. Once a defendant raises a concrete licensee-marking challenge, a patent owner can no longer assume that a general assurance of compliance will carry the case into discovery. That changes enforcement planning even for an owner who would ultimately win the marking question.

How today's settlement burdens tomorrow's lawsuit

Settling an infringement case does not put the resulting license beyond the reach of the marking statute. The danger is specific. It arises when the agreement authorizes products that actually practice the patent, those products go out unmarked, and nothing (no reasonable compliance effort, no timely actual notice) supports the damages period the owner wants. A license is not an admission that every product the licensee sells practices every licensed patent.^[1][2][5]

Having accused a company of infringement makes it awkward to argue later that its products never practiced the patent. Awkward is not fatal. An accusation is not a judicial finding of coverage, and VDPP expressly rested on the owner's continued infringement position and the particular agreements before it. Examine those facts; do not assume that a signature on any settlement forfeits recovery.

An express no-marking clause is a red flag; read it today. Silence deserves review too, though the two are not the same, and neither answers by itself whether there were patented articles to mark. The reasonable-efforts inquiry is still open. What no one yet knows is how much cooperation an owner must extract from an independent company, particularly when the deal gives the owner little leverage.^[1][5]

Scope matters too. A release limited to past conduct is a different animal from permission to keep making and selling patented products. Ortiz flagged that distinction as an argument the owner never made. The label on the document, whether release, covenant, or dismissal, settles nothing; what counts is the rights it actually grants.^[4]

That uncertainty shows up at the negotiating table. A settling defendant has every reason to resist identifying products, committing to mark, or promising future cooperation. The owner must decide what those protections are worth in a lawsuit against someone else, years from now. A term that looked peripheral to the settlement price can turn out to control the value of the patent.

The bargaining problem exists before the next case does. Push for stronger marking terms and you may give up money or take on administrative work that never ends. Accept weaker terms and you keep today's deal at the price of tomorrow's damages. Neither choice is free. VDPP simply makes it harder to put off the calculation.

Why the big-company cross-license is the real exposure

The scale of the problem is visible in the companies' own press releases. In January 2014, Samsung and Google announced a global cross-license spanning a broad range of technologies and covering existing patents and patents filed over the next ten years. In May 2021, Ericsson and Samsung announced a global agreement covering cellular technologies and sales of network infrastructure and handsets; the settlement ended their pending patent disputes.^[6][7]

To be clear, those announcements show breadth, not a marking failure. They say nothing about the agreements' marking terms, which products practice which claims, or what either company did to secure marking. Nothing about either company's recoverable damages follows from a press release. What the releases show is the shape of the deals to which this legal risk now attaches.

A broad cross-license buys peace without answering which product practices which claim. A later lawsuit can reopen every one of those questions, for a different purpose. Under Arctic Cat's framework, the accused infringer needs only to identify specific unmarked products it believes practice the asserted patent. That showing is modest. The burden then shifts to the patentee to prove compliance or to prove that the identified products do not practice the patent. A press release announcing a cross-license is not that product-specific proof.^[5]

The irony is that the deal was supposed to end this kind of work. The Justice Department and Federal Trade Commission have long recognized that cross-licensing reduces transaction costs, clears blocking positions, and avoids infringement litigation. A marking dispute drags the owner back into technical investigation and negotiation with a company it already paid to leave alone. That is the friction licensing teams should start measuring.^[8]

Where the money goes: the work and the lost bargaining power

Put numbers on it, as an illustration rather than a forecast. Assume a valid, infringed patent, a properly apportioned royalty base of $1 billion a year, and a 1 percent royalty. Five years is $50 million before other adjustments. If a marking failure bars that entire pre-notice period, the $50 million is gone. The sales happened. The infringement happened. The recovery did not.

Section 286 sets a six-year outer limit on pre-filing recovery; it promises nothing. Section 287 can shrink that window to zero. The illustration assumes no effective marking or qualifying notice during the five years and that every other requirement for recovery is met. Real losses turn on the patent, the products, the dates, and the evidence.^[2]

Even when the damages survive, proving it costs money. Counsel will need old agreements, product samples, technical records, and evidence of marking over time. Engineers will have to compare a partner's products against the asserted claims. Acquisitions and product changes muddy the trail. And because the partner controls the information, a simple enforcement decision can become a long investigation of someone else's business.

The uncertainty reaches valuation too. A prospective licensee will discount the threat of historical damages; a portfolio buyer will demand proof before paying for it. None of this is an automatic write-down. A portfolio's value still depends on remaining term, claim scope, validity, infringement, future royalties, and other remedies. But a damages period you cannot defend is a weaker chip at the table.

Repair is uneven and often incomplete. An old counterparty has little reason to accept new marking or cooperation duties without being paid for them. Fresh compliance can protect later periods where the law allows; it cannot create notice that did not exist years ago. A company may pay to investigate the problem, pay to fix its agreements, and still lose the historical recovery. For a heavily licensed portfolio with real marking gaps, that is a slow economic bleed.

Claim selection can change the answer, but not reliably. Crown Packaging held that section 287(a) does not apply when only method claims are asserted. But method claims are no universal escape hatch. They must support a viable infringement case, and Rembrandt rejected an attempt to erase past marking consequences by disclaiming the claim the licensed products practiced. Dropping claims mid-course is no substitute for examining the history.^[9]

Confidentiality concerns add expense, not inevitable public exposure. Relevant agreements may become discoverable, but VDPP does not require every license or royalty term to appear in a public complaint. Protective orders can limit who sees confidential commercial terms, and sealing a filing requires its own justification under the court's rules. A rival's outside counsel reviewing protected material under a protective order is a different thing from the rival's business team reading it.^[10]

Nor can a sophisticated defendant's knowledge substitute for notice. Notice has a technical meaning. A defendant may know about the patent, and even know it infringes, without ever receiving the affirmative notice section 287 requires. Amsted demands a specific charge of infringement identifying the accused product. Filing suit counts as notice, but it recovers nothing from the period already lost.^[2][11]

Expired patents raise the stakes further. With no term left, there is no future infringement to recover for; the past is all there is. An expired patent is still enforceable, and historical damages remain available where marking was satisfied, never required, or replaced by timely actual notice. Nor does every large company come out behind. The company that loses as a plaintiff may win as a defendant. Whether the industry nets out ahead is an empirical question. The exposure on your own docket is not.

The legal question that deserves a better answer

The court's public-notice rationale is serious and deserves to be taken seriously. Unmarked patented products leave the public without the notice Congress intended, and a private license does not erase that concern. The harder question is how an owner is supposed to establish compliance when another company controls the products and the information. A workable rule must preserve notice without demanding a full forensic audit as the price of filing suit.

Wine Railway does not hand a nonproducing owner a blanket exemption once its licensees are making products. In that 1936 case no one, not the owner and not anyone acting with its consent, had made or sold an article under the patent. Its reasoning carries weight when there truly is no authorized patented article to mark. It does not automatically excuse licensed manufacturing. And the duty to plead and prove notice was recognized long before VDPP, including in Dunlap.^[12][13]

The procedural question is still live. How much must a complaint say before the defendant has identified the products it claims required marking? Rule 9(c) allows conditions precedent to be pleaded generally, but whether that rule governs this damages prerequisite has to be argued, not assumed. Courts should also keep a deficient complaint separate from a disputed evidentiary record. Rule 12(d) governs conversion when outside material comes in, subject to the recognized pleading-stage exceptions.^[5][10]

VDPP should be read against its own record. A court applying it to a portfolio cross-license should ask what was authorized, which products practiced the patent, over what period, and what efforts the owner made, rather than treat the existence of a deal as an answer to all four. The case for further review is strongest at exactly that boundary: how to preserve public notice without licensing an unbounded investigation into a counterparty's technology.

What corporate America should do next

Start with the patents and damages periods that matter commercially. Identify the licensees and products, read the agreements, and preserve evidence of marking and notice. A targeted audit beats a blanket assumption that every patent in every cross-license carries the same defect. And enforcement valuations should say which historical revenues are actually recoverable on the facts, rather than book the six-year lookback as cash.

Future agreements should spell out workable marking and cooperation duties: who identifies products, who keeps the virtual-marking page current, who keeps the compliance records. Document every request and every response. A refused request may be useful evidence, but neither VDPP nor Ortiz guarantees that one letter satisfies the reasonable-efforts standard. Marking your own products does not solve an independent licensee's failure to mark.^[1][4][5]

In litigation, plead the basis for recovery you can support and keep the distinctions sharp. Plead facts showing compliance, inapplicability, or actual notice. If the agreement is a limited release, say so and explain why. If dismissal is threatened, propose the specific factual amendment that cures the defect and address Rule 15 head on. Disagreeing with the marking rule in the abstract is no substitute for a usable record.^[10]

Appellate strategy has to be just as concrete. The right case can seek en banc review under Federal Rule of Appellate Procedure 40, Supreme Court review, or a decision confining VDPP to its facts. Congress can consider a statutory solution. The Ortiz certiorari petition is no longer a vehicle; the Supreme Court denied it on June 29, 2026. That denial said nothing about the merits of the marking debate.^[14][15]

Operating companies can supply what the debate so far lacks: real cooperation provisions, the actual cost of identifying licensed products, the practical limits of virtual marking, and the effect on settlement prices. A safe harbor for documented reasonable efforts would need a credible public-notice mechanism behind it. It should be pitched as reform, not as an exemption the law already grants.

Large companies should not file VDPP under someone else's patent-assertion problem. A rule written against one small owner now governs the agreements that keep much larger competitors out of court. When buying peace with one company weakens your recovery against another, that cost belongs in the boardroom. The next defendant may already be reading your license file.

Disclosure: This article develops memoranda prepared for William P. Ramey III, who represented VDPP and Ortiz in the litigation discussed here.^[1][14]

Sources

[1] VDPP, LLC v. Volkswagen Grp. of Am., Inc., 186 F.4th 1372, 1376–77 (Fed. Cir. 2026) (precedential). The marking analysis concerns the specific licenses and proposed amendment; the fee ruling also rests on other conduct. Official opinion

[2] 35 U.S.C. §§ 286, 287(a). Section 286 limits the lookback; section 287(a) addresses marking, actual notice, and the filing of suit as notice. Section 286 | Section 287

[3] Arctic Cat Inc. v. Bombardier Recreational Products Inc., 950 F.3d 860, 863–67 (Fed. Cir. 2020). Cessation of unmarked sales and willfulness did not establish the notice necessary for pre-complaint damages. Opinion

[4] Ortiz & Associates Consulting, LLC v. Vizio, Inc., No. 2024-1783, slip op. at 3, 6 & n.2, 7–9 & n.4 (Fed. Cir. Dec. 17, 2025) (nonprecedential). The appeal concerned fees; the panel identified arguments and allegations the owner had not made. Official opinion

[5] Arctic Cat Inc. v. Bombardier Recreational Products Inc., 876 F.3d 1350, 1366–69 (Fed. Cir. 2017). The opinion states the product-identification burden and discusses the licensee reasonable-efforts rule, citing Maxwell v. J. Baker, Inc., 86 F.3d 1098, 1111–12 (Fed. Cir. 1996). Opinion

[6] Samsung, “Samsung and Google Sign Global Patent License Agreement” (Jan. 27, 2014). The announcement establishes the stated scope of the transaction, not its marking terms or product coverage under any particular patent. Company announcement

[7] Ericsson, “Ericsson and Samsung sign global patent license agreement” (May 7, 2021). The announcement identifies covered technologies, products, and the settlement of litigation; it does not establish a marking deficiency. Company announcement

[8] U.S. Department of Justice & Federal Trade Commission, Antitrust Guidelines for the Licensing of Intellectual Property § 5.5, at 30 (Jan. 12, 2017). The agencies describe potential benefits of cross-licensing and pooling, while also recognizing possible anticompetitive effects. Guidelines

[9] Crown Packaging Technology, Inc. v. Rexam Beverage Can Co., 559 F.3d 1308, 1316–17 (Fed. Cir. 2009); Rembrandt Wireless Technologies, LP v. Samsung Electronics Co., 853 F.3d 1370, 1382–84 (Fed. Cir. 2017). Method-only assertion and a later statutory disclaimer present distinct questions. Crown Packaging | Rembrandt

[10] Fed. R. Civ. P. 9(c), 12(b)(6), 12(d), 15(a)(2), 26(c)(1)(G). The rules govern general allegations of conditions precedent, dismissal, outside materials, amendment, and protection of confidential commercial information. A protective order does not itself authorize sealing a court filing. Official Civil Rules as amended Dec. 1, 2025

[11] Amsted Industries Inc. v. Buckeye Steel Castings Co., 24 F.3d 178, 187 (Fed. Cir. 1994). Actual notice requires an affirmative, specific infringement charge; knowledge alone is insufficient. Opinion

[12] Wine Railway Appliance Co. v. Enterprise Railway Equipment Co., 297 U.S. 387, 393–98 (1936). The stated facts include the absence of manufacture or sale by the owner or anyone acting with its consent. Opinion

[13] Dunlap v. Schofield, 152 U.S. 244, 247–48 (1894). The Court assigned the duty of allegation and burden of proof concerning notice to the patentee in the circumstances addressed there. Opinion

[14] Supreme Court docket No. 25-1326, Ortiz & Associates Consulting, LLC v. Vizio, Inc. The June 29, 2026 entry records denial of certiorari. Docket checked Oct. 15, 2026. Official docket

[15] Fed. R. App. P. 40 (panel rehearing and en banc determination); Supreme Court Rules 10, 13 (discretionary review and filing time). Availability and deadlines depend on the actual judgment, rehearing orders, and any extension; no case-specific VDPP filing deadline is asserted here. Rule 40 | Supreme Court Rules effective Mar. 16, 2026

Ramey LLP is a Texas-based intellectual property law firm dedicated to representing small patent owners, startups, and independent inventors in disputes against larger corporations.