Federal Circuit Vacates Infringement, Damages Against Apple

Somewhere in East Texas, a jury checked a box that basically said, “Yep, Apple infringed something,” and nine figures promptly started packing their bags. Then the U.S. Court of Appeals for the Federal Circuit showed up like the group-chat friend who actually reads the fine print, and said: “About that… no.”

In Optis Cellular Technology, LLC v. Apple Inc., the Federal Circuit vacated (legal-speak for “hit undo”) both the infringement finding and the $300 million damages judgment, sending the case back for a new trial on infringement and damages. The opinion is a master class in how appellate courts can topple a massive verdict without ever arguing that smartphones don’t do 4G.

The Case in One Espresso Shot

Optis and related entities asserted five LTE standard-essential patents (SEPs) against Apple devices (think iPhone, iPad, Apple Watchanything that likes fast cellular connectivity). A first trial produced a $506.2 million award. The district court later ordered a new damages trial, largely because the first jury didn’t hear certain evidence about Optis’s FRAND licensing obligations (Fair, Reasonable, and Non-Discriminatory terms that often travel with SEPs like luggage fees). A second jury returned a $300 million lump-sum award.

On appeal, the Federal Circuit didn’t merely tweak the number. It tossed the infringement and damages judgments entirely and remanded for proceedings consistent with its opinion which is the appellate equivalent of dumping your LEGO set onto the floor and telling you to rebuild it “the right way.”

How We Got Here: Two Trials, One Very Expensive “ANY”

The story turns on a deceptively small word: ANY. In the first trial, the verdict form asked the jury a single broad infringement questionwhether Apple infringed “ANY of the asserted claims.” That sounds efficient… until you remember the case involved five different patents.

A quick timeline

• 2019: Optis sues Apple in the Eastern District of Texas over LTE SEPs.
• 2020: First trial ends with a roughly $506.2 million verdict for Optis.
• 2021: District court orders a new trial on damages; the second jury awards $300 million as a lump sum.
• 2025: Federal Circuit vacates infringement and damages judgments; case is sent back for a new trial on both.

The key point: the second jury’s damages award rested on an underlying infringement finding that (according to the Federal Circuit) wasn’t properly “built” in the first place. If infringement is the foundation, damages are the fancy penthouse. And the court decided the foundation had some… structural issues.

Why the Federal Circuit Hit Undo: Four Big Problems

The appellate opinion isn’t a one-issue thriller. It’s a multi-plot season finale. Here are the big themes that mattered most.

1) Jury Unanimity and the “One-Question” Verdict Form

In civil jury trials, unanimity is the default unless the parties stipulate otherwise. The Federal Circuit emphasized that each asserted patent is its own legal claim. That means a jury can’t be asked a single global question if the result allows jurors to disagree on which patent was infringed while still producing a “yes” answer.

Here’s the practical problem. Imagine a jury of eight: four jurors think Patent A is infringed, and four think Patent B is infringed. Nobody gets eight votes on any single patentbut under a broad “ANY of the asserted claims” question, the group can still produce a unified “yes.” That’s not a unanimous verdict on any particular patent claim; it’s a unanimous verdict that something somewhere happened.

The Federal Circuit’s message to trial courts is blunt: if you want a verdict that survives appeal, separate the questions at least patent-by-patent. Efficiency is great, but not at the expense of a party’s right to a proper jury verdict.

2) Damages Evidence: When a Settlement Number Becomes a Megaphone

Patent damages often revolve around “comparable licenses”other deals that can help estimate a reasonable royalty. The problem is that not every license is actually comparable, especially when a number is tied to global litigation peace rather than a narrow license to a handful of patents.

In this case, evidence about Apple’s prior settlement with Qualcomm (from a different dispute) was used in the damages story. The Federal Circuit found its probative value minimal because the settlement covered a much broader set of rights and resolved wide-ranging global litigation. Meanwhile, the risk of unfair prejudice was highlarge settlement figures can anchor a jury’s expectations and distort how jurors think about value.

Put simply: if jurors repeatedly hear a giant number, it can become the gravitational center of the damages universeeven if it’s not tied to the actual patents being litigated. The Federal Circuit concluded the settlement evidence should have been excluded under Federal Rule of Evidence 403.

3) Patent Eligibility Under § 101: When “The Innovation” Is Mostly Math

Standard-essential patents often sound technical (because they are), but § 101 disputes can still ariseespecially when the “advance” looks like a mathematical formula or algorithm described at a high level.

The Federal Circuit reversed the district court’s determination that certain claims (including claims 6 and 7 of one asserted patent) were not directed to an abstract idea. The appellate court concluded those claims were directed to the abstract idea of using a mathematical formula. Because the district court had stopped after step one of the Alice framework, the Federal Circuit sent the issue back for a proper step-two analysis (the “inventive concept” inquiry).

Translation for non-patent-law humans: even in a cellular-network context, “we do math to do the thing” may not be enough. The claims must tie the idea to a concrete technical improvement in a way that isn’t just results-oriented or conventional implementation.

4) Means-Plus-Function (§ 112): “Unit” Isn’t Always a Get-Out-of-Structure-Free Card

Drafters sometimes use terms like “module,” “device,” or “unit” hoping they’ll sound structural while staying flexible. Courts, however, have learned that these can be nonce termsplaceholders that describe function without specifying structure.

The Federal Circuit held that a claim term like “selecting unit” invoked means-plus-function treatment under § 112 ¶ 6, rejecting the notion that “unit” by itself necessarily connotes sufficient structure. The consequence is significant: once a term is treated as means-plus-function, it must be linked to adequate corresponding structure in the specification, or the claim can run into indefiniteness trouble. The court remanded for the district court to handle the remaining step of that analysis.

Why This Decision Matters Beyond Apple vs. Optis

You don’t need to own a smartphone (or a patent portfolio the size of a phone book) to see the ripple effects. This opinion matters because it targets pressure points that show up repeatedly in high-stakes patent casesespecially SEP disputes:

• Verdict engineering: The way questions are framed can decide whether a verdict lives long enough to be collected.
• Comparable licenses: Courts are increasingly skeptical of “comparable” deals that aren’t truly comparable in scope, geography, or dispute posture.
• SEP / FRAND context: Juries can hear about FRAND obligations, and the presence (or absence) of that evidence can drive retrial decisions.
• Drafting discipline: If a claim leans on math or functional language without structural detail, it may invite § 101 or § 112 problemseven years later.

There’s also a strategic reality here: SEP disputes are frequently global. While U.S. litigation focuses on patent-by-patent infringement and U.S. damages rules, other jurisdictions may approach FRAND and licensing differently. When a major U.S. verdict gets vacated, it can change negotiation leverage, settlement posture, and the storytelling in parallel proceedings.

Practical Takeaways for Companies and Trial Teams

For litigators: build verdict forms like you’ll have to defend them on appeal

• Use separate infringement questions at least by patent (and sometimes by claim) to protect unanimity and clarity.
• Match the structure across issues: if invalidity is broken out by claim, consider whether infringement should be similarly granular.
• Plan for retrial scenarios: if damages are retried without liability, ensure the record makes clear exactly what the liability finding was.

For in-house counsel: treat “big numbers” as risk multipliers

• A large settlement in a different matter can become a trial exhibit in the next matterunless carefully handled through motions in limine and evidentiary objections.
• Licensing strategy is litigation strategy. If your licenses aren’t cleanly comparable, be ready for a fight over what (if anything) they prove.

For patent owners and prosecutors: structure beats vibes

• If a claim relies on an equation, spell out the technical improvement and how the implementation avoids conventionality.
• If using “unit/module/device,” make sure the claim language and specification provide concrete structureor be prepared for § 112 scrutiny.

What Happened Next: The “Third Trial” Chapter

Remands don’t end cases; they restart them. After the Federal Circuit sent Optis and Apple back for a new trial, the dispute continued. In a later jury trial in the Eastern District of Texas, Apple reported a win on Optis’s infringement allegationsshowing how dramatically the posture of a case can swing after an appellate reset.

That doesn’t mean the story is over. In SEP litigation, post-trial motions and appeals are practically a second sport. But it does underline the real-world impact of the Federal Circuit’s decision: vacating a verdict doesn’t merely change a number on a spreadsheetit can change the entire direction of the litigation.

Experience Corner: What It Feels Like When a $300 Million Verdict Gets “Vacated” (About )

If you’ve never been close to a major patent trial, here’s the part people don’t tell you: the biggest drama often isn’t the technology. It’s the choreography. Not the kind with jazz handsmore the kind where every witness, exhibit, and jury instruction has to land on a single storyline that still makes sense a year later to three judges reading cold transcripts.

Multi-patent cases are especially prone to “narrative shortcuts.” Everyone wants to simplify: one chart, one theme, one clean question for the verdict form. And in the moment, it feels reasonable. Jurors are human. If you throw five patents, dozens of claims, and a mountain of LTE jargon at them, their eyes can glaze over faster than a donut. So the temptation is to ask a broad question“Did the defendant infringe any of the asserted claims?”and move on.

But appellate courts don’t grade on a curve. When you zoom out, that single question can hide a fatal ambiguity: What exactly did the jury unanimously decide? Trial teams often discover too late that “clarity for jurors” and “clarity for appeal” aren’t always the same thing. On appeal, the judges aren’t watching witnesses; they’re auditing logic. And logic hates mystery. If different jurors could have reached the same “yes” answer for different reasons across different patents, the verdict becomes less like a unanimous decision and more like a group project where nobody remembers who wrote the conclusion.

Damages brings its own version of chaos. In a big case, everyone knows that one giant number can shape the room. Mention a massive settlement figureeven from a different dispute and it can become a mental anchor. You can practically feel the temperature change when a jury hears a number with lots of zeros. The problem is that anchoring is sticky: even if an expert insists the number is “just a reference point,” jurors are still left thinking, “If that deal was huge, this one must be huge too.” Trial lawyers spend enormous energy trying to keep the damages conversation tied to the patents actually on trial, not the gravitational pull of unrelated headline figures.

Then there’s the drafting-and-claim-language side of the experience. Engineers describe solutions. Patent claims describe boundaries. Those are different jobs, and sometimes the claim language ends up sounding like: “a unit configured to do the thing.” In litigation, that’s like wearing a suit made of legal Velcroeverything sticks to it, including means-plus-function arguments. When a court decides “unit” is just a placeholder for a function, you’re suddenly deep in the specification hunting for concrete structure like it’s a scavenger hunt with real consequences.

The big lessonespecially in a case like this oneis that “winning at trial” and “keeping the win” are different disciplines. A verdict that looks simple can be dangerously fragile. The teams that survive appeals are the ones who treat jury forms, evidentiary choices, and claim language as engineering problems: define the parts, test for ambiguity, and assume someone will try to break itbecause they will.

Conclusion

The Federal Circuit’s decision to vacate infringement and damages against Apple isn’t just another appellate reversalit’s a playbook for how big patent verdicts fail: a verdict form that blurs separate patent claims, damages evidence that risks unfair prejudice, and claim issues that invite § 101 and § 112 scrutiny.

Whether you’re a patent owner, an accused infringer, or a business leader watching SEP litigation shape licensing strategy, the takeaway is simple: precision wins. Not just in the technologyalso in the questions you ask a jury, the evidence you put in front of them, and the words you choose when you draft a claim. In modern patent litigation, the smallest word on the verdict form can be the most expensive.