Narrowed Scope of IPR Estoppel from Federal Circuit

Not legal advice. Patent litigation is complicated, jurisdiction-specific, and occasionally powered by coffee and deadline panic.

Inter partes review (IPR) was supposed to be the “fast, focused” way to test patent validity at the Patent Trial and Appeal Board (PTAB).
Then came IPR estoppelthe rule that tells a challenger, “Okay, you took your shot. You don’t get infinite do-overs.”
For years, lawyers argued about how big that “no do-overs” zone really is.
In 2025, the U.S. Court of Appeals for the Federal Circuit drew a sharper line, and it matters a lot for how both sides plan a case.

This article explains what changed, why it changed, and what it means in the real worldespecially after the Federal Circuit’s decision in
Ingenico Inc. v. IOENGINE, LLC, which many commentators describe as narrowing the practical reach of IPR estoppel in district court.

IPR Estoppel in Plain English

What is IPR?

IPR is an administrative proceeding where a petitioner asks the PTAB to cancel patent claims as unpatentable. The key constraint:
IPR challenges are limited to invalidity theories under 35 U.S.C. §§ 102 and 103, and only when based on
patents and printed publications. In other words, IPR is a document-based cage match. No “we sold this product before the patent,”
no “we used this in public,” and no “this claim is indefinite” (those live elsewhere).

What is estoppel?

Estoppel is the “you already had your chance” rule. Under 35 U.S.C. § 315(e), if an IPR results in a final written decision,
the petitioner (and related parties) may be barred from raising certain invalidity positions latereither:

• Before the USPTO (Section 315(e)(1)), or
• In district court or the ITC (Section 315(e)(2)).

The policy idea is simple: IPR is meant to streamline disputes, not create an endless carousel of “same patent, different day.”
But the devil lives in a single word: ground.

The Big Question: What Counts as a “Ground”?

Section 315(e)(2) bars a petitioner from asserting in court that a claim is invalid on any “ground” the petitioner raised
or reasonably could have raised during the IPR. Courts split over what that meant in practice.

Two competing views (and lots of billable hours)

• Broad view: Estoppel can block not just the exact petition grounds, but also later invalidity arguments that rely on
  “the same stuff” (or substantially similar prior art), even if repackagedespecially if a petitioner could have used that material in the IPR.
• Narrow view: Estoppel blocks only what an IPR is legally capable of decidingmeaning only certain document-based
  invalidity theoriesand does not reach other theories like public use or on-sale activity, even if related documents exist.

If you’re thinking, “Waitcan’t a product be basically the same as a printed manual?” Congratulations, you’re now ready for the next season
of Patent Litigation: The Miniseries.

The Federal Circuit’s 2025 Line-Drawing: Ingenico v. IOENGINE

In Ingenico Inc. v. IOENGINE, LLC (May 7, 2025), the Federal Circuit tackled the meaning of “ground” in Section 315(e)(2)
and resolved a district-court split. The court’s bottom line is the part litigators immediately highlighted in red, bold, and possibly a panic font:

IPR estoppel applies only to a petitioner’s assertions in district court that a claim is invalid under §§ 102 or 103
because it was patented or described in a printed publication (or would have been obvious only on the basis of prior art patents or printed publications).
It does not bar invalidity theories that could not be raised in IPR, such as known or used by others, on sale, or in public use.

Key holding #1: “Ground” means the legal theorynot the evidence

The court emphasized that Congress could have written the statute to estop a petitioner from asserting “any prior art” it raised or could have raised.
But Congress did not. It used “ground,” and the Federal Circuit read “ground” in context across the Patent Act to mean a theory of invalidity.

Translation: estoppel is aimed at the kind of invalidity argument (printed-publication anticipation/obviousness) rather than a universal ban on
using certain references ever again. Evidence can sometimes travel, even when a particular theory cannot.

Key holding #2: You can reuse IPR documents as evidence for non-IPR theories

The decision clarifies that a petitioner is not automatically forbidden from using the same patents or printed publications seen in an IPR.
Those documents may still show up in district court as evidence supporting a different theory that IPR does not allow.

Example: If a defendant argues the invention was on sale or in public use before the patent’s critical date,
manuals, readme files, datasheets, and other documents can help prove what the product was and what it didwithout converting the argument into a
printed-publication-only “ground.” The theory is “on sale/public use,” and the documents are corroboration.

Why the court said this makes sense

The Federal Circuit leaned heavily on statutory structure:

• Section 311(b): IPR is limited to grounds under §§ 102 and 103, and only on the basis of patents and printed publications.
• Section 312(a)(3): A petition must identify “grounds” and provide evidence (including copies of patents/printed publications) supporting those grounds,
  suggesting a distinction between ground and evidence.
• Comparison to post-grant review (PGR): PGR can raise broader invalidity theories (including certain § 112 issues), and PGR’s estoppel provision is worded similarly,
  reinforcing that “ground” tracks what the proceeding can legally address.

Bottom line: if IPR was designed as a streamlined, document-focused process, estoppel should not magically expand it into a
“you can never argue anything related to this topic again” super-estoppel.

Where This Fits with Earlier Federal Circuit Estoppel Cases

Caltech and the “reasonably could have raised” bite

In California Institute of Technology v. Broadcom (2022), the Federal Circuit discussed estoppel principles and reinforced that estoppel can reach
invalidity arguments a petitioner reasonably could have raised in IPRan idea that can be broad when you’re talking about
additional patents and printed publications that a petitioner could have found and included.

That “reasonably could have raised” phrase is still a big deal after Ingenico. The narrowing is not “estoppel is gone.”
It’s “estoppel is tied to what IPR actually allows.”

Ironburg and the “skilled searcher” standard

In Ironburg Inventions Ltd. v. Valve Corp. (2023), the Federal Circuit endorsed a practical approach for deciding whether an invalidity ground
“reasonably could have been raised” in IPR: would a skilled searcher conducting a diligent search reasonably have been expected to discover it?
That standard influences whether late-found printed-publication grounds are barred after a final written decision.

Put differently: if you lose an IPR and then “discover” a conveniently devastating journal article that any competent search would have found before filing,
you may have a hard time avoiding estoppel on that printed-publication anticipation/obviousness theory. Ingenico doesn’t change that;
it changes what happens when your later theory is not the printed-publication theory at all.

Concrete Examples: What’s Blocked vs. What Survives

Scenario A: The classic printed-publication 102/103 attack

You file an IPR arguing Claim 1 is obvious over Patent X + Article Y. You get a final written decision.
Later in district court, you try to argue Claim 1 is obvious over Patent X + Article Z (which you could have found with a diligent search).
That’s a textbook estoppel fight, and Ironburg-style reasoning may matter a lot.

Scenario B: The “product prior art” pivot

You also have evidence that a product embodying the claimed features was sold publicly before the patent’s critical date.
You want to argue invalidity based on on-sale or public use. You introduce the product documentation (manuals, readme files, spec sheets)
and testimony to show what was sold and when.

After Ingenico, that theory is not automatically swept into IPR estoppel simply because some of the documents could have been used as
printed publications in an IPR. The “ground” is different, and IPR does not permit on-sale/public-use grounds.

Scenario C: The “don’t relabel it and hope nobody notices” problem

Courts can still scrutinize whether a party is genuinely asserting a non-IPR ground or just dressing up a printed-publication theory in a new outfit.
If your “public use” theory has no real public-use facts and is just “this brochure existed,” that starts to look like a printed-publication ground again.
Good lawyering requires real evidencedates, witnesses, sales records, product samples, and authenticationbecause public-use/on-sale cases can be fact-heavy.

Strategy Impacts for Petitioners and Patent Owners

For defendants/petitioners: plan a two-track invalidity playbook

• Don’t treat IPR as your whole defense. IPR is powerful, but it’s document-limited. If you may need on-sale/public-use defenses,
  start building that factual record early.
• Separate “theory” from “evidence” on purpose. After Ingenico, you may be able to use the same documents as evidence
  for non-IPR theories. Your invalidity contentions should make the theory explicit.
• Do the search like you mean it. If you intend to rely on printed publications under §§ 102/103, assume the other side will argue you
  “reasonably could have raised” additional references in the IPR. Build a record of diligence.
• Watch timing and parallel USPTO options. The Federal Circuit’s later decision in In re Gesture Technology Partners (Dec. 1, 2025)
  highlights that IPR estoppel under § 315(e)(1) may not automatically shut down an ongoing ex parte reexamination once it’s underway, because the USPTOnot the requestermaintains it.
  That can influence how parties think about parallel proceedings.

For patent owners: expect more “non-IPR” defenses after an IPR loss

• Push for early disclosure of on-sale/public-use theories. These defenses can be broad, messy, and expensive. Case management matters.
• Attack proof problems. Public use and on-sale defenses often rise or fall on corroboration, documentation, and credible dates.
  If the story is fuzzy, focus on admissibility and reliability.
• Use estoppel surgically. Estoppel remains very real for printed-publication § 102/103 theories, especially when late-found references look
  like they should have been in the petition.

What to Watch Next

Even with Ingenico drawing a clearer line, litigation will keep probing the edges:

• How courts police “evidence reuse.” When does evidence effectively become the “ground”?
  If the theory is nominally “on sale,” but the entire case hinges only on a printed document, expect disputes.
• The scope of “reasonably could have raised.” The skilled-searcher standard matters, and parties will keep fighting about
  what searches were reasonable, what databases count, and what diligence looks like for the technology at issue.
• Different forums, different pressure points. District courts, the ITC, and the USPTO apply estoppel provisions in different procedural postures,
  which can affect leverage and timing.

Conclusion

The Federal Circuit’s recent treatment of IPR estoppel is best understood as a “scope correction,” not an “estoppel delete button.”
After Ingenico v. IOENGINE, estoppel is tied more tightly to what IPR can actually adjudicate: printed-publication and patent-based
anticipation/obviousness theories under §§ 102 and 103. That clarification can leave room for defendants to litigate
on-sale/public-use/known-or-used theories in district court, even while remaining estopped from document-only theories that were or
reasonably could have been raised in the IPR.

If you’re litigating patents, the practical message is simple: treat IPR as a high-powered tool, not a whole toolbox.
And if you’re a patent owner, assume the toolbox gets biggernot smallerafter the IPR ends.

“Experience” Section: What This Looks Like in Practice (500+ Words)

Here’s a truth practitioners will tell you with the same tired smile they reserve for “quick questions” at 6:59 p.m.:
nobody wants to litigate IPR estoppel. People litigate IPR estoppel because the alternative is worselike explaining to a client why
the best invalidity argument is now sitting behind a velvet rope labeled “should’ve put that in the petition.”

Before Ingenico, a common anxiety went like this: “If we file an IPR and lose, does that mean the district court case becomes a
one-defense sitcom with no new episodes?” The fear wasn’t imaginary. Some courts were willing to apply estoppel broadly, especially when later
invalidity positions looked suspiciously like a remix of the IPR petitionwith the same beat, same chorus, and the same awkward dance move.

After Ingenico, the conversation in many litigation teams shifts from “Are we barred?” to “What theory are we actually asserting?”
That sounds obvious, but in day-to-day practice it’s not. Teams frequently start with a pile of facts (a product existed, a brochure existed,
an engineer remembers a demo, an invoice exists somewhere in a backup server that nobody has opened since the Obama administration) and then
try to back into a legal theory. The Federal Circuit’s “ground vs. evidence” framing encourages the opposite: pick the legal theory deliberately,
then decide what evidence truly supports it.

One very common “experience” pattern: document-based IPR arguments and fact-based “real-world” defenses tend to live in different departments of the same
company. The patents/printed publications search often comes from counsel and search professionals. The on-sale/public-use story often comes from
product teams, sales operations, and long-retired engineers who still have a garage full of prototypes. (Sometimes those prototypes are the best witness.
Sometimes they are, unfortunately, a box of mystery parts and a sticky note that says “rev B???”)

So what do teams do now? They build parallel tracks early. While the IPR petition is being assembled with its strict page limits and citation rules,
a separate “fact track” starts collecting product samples, release notes, internal emails, archived web pages, dated manuals, purchase orders, and anything that can
establish a timeline. The goal is not to “escape estoppel” by clever labeling; the goal is to ensure that if a non-IPR theory is legitimately available,
it is supported by real evidence and can survive motions challenging authenticity, hearsay, or gaps in corroboration.

Another practical lesson people learn quickly: a non-IPR theory can be harder to prove than a printed-publication theory. Public use and on-sale defenses may require
credible witnesses, clean business records, and a coherent story about what exactly was disclosed or sold, to whom, and when.
That means the post-Ingenico “extra room” in district court is not free spaceit’s space you still have to furnish.

Finally, Ingenico changes negotiation dynamics. When one side previously believed estoppel would “shrink” the case after an IPR, they might have expected
settlement leverage to spike. Now, experienced litigators assume the case may re-expand through other theories, and they plan settlement strategy accordingly.
The humor here is that patent litigation remains a game of chessexcept the pieces are legal theories, the clock is discovery, and someone keeps moving the board
because a new precedential decision just dropped.