The Reexamination End-Run: When Second Bites at the Apple Become Strategy

A March 27, 2026, petition filed by Security First Innovations (SFI) does more than challenge a single reexamination proceeding—it shines a spotlight on a structural vulnerability in how post-grant review is functioning in practice. At its core, the filing argues that the U.S. Patent and Trademark Office (USPTO) is allowing what amounts to a procedural end-run around the statute, which is supposed to streamline post-grant challenges and lead to estoppel if the patent owner prevails.

This petition highlights a problem presented in a growing number of cases. An inter partes review (IPR) fails and challengers simply repackage the filing as an ex parte reexamination request. If this practice is not stopped by the USPTO, all the Patent Trial and Appeal Board (PTAB) reforms put in place by Director John Squires will have been for naught.

The Office is Encouraging a System of Duplication

The facts in this case are not subtle. After IBM failed to secure institution of an IPR, it filed an ex parte reexamination request just days later. IBM did not materially revise its invalidity case; it simply re-labeled it. SFI explains that IBM “essentially replaced the caption of its failed IPR petition and resubmitted the same document” as a reexamination request. The overlap is complete: the same claims are challenged using the same references—Dickinson, Hardjono, and Moulton—and the same combinations and arguments. IBM even acknowledges that the grounds are “in substance, identical.”

This is not a situation where new evidence or a refined theory is being presented by a patent challenger. It is duplication, pure and simple. And that duplication matters because Congress anticipated precisely this type of behavior when it enacted 35 U.S.C. § 325(d). The statute gives the USPTO discretion to deny review when “the same or substantially the same prior art or arguments” have already been presented to the Office. The policy objective was straightforward: challengers get one meaningful opportunity to invalidate a patent, not multiple attempts across different procedural vehicles.

The current system is drifting away from the one-challenge principle because it treats IPR and ex parte reexamination as distinct channels, even when the substance is identical. And, if we are honest, IPR and ex parte reexamination are not distinct. From the patent owner’s perspective, both procedures seek to take patent claims away. So, while we intellectually argue about different standards and different processes, at the end of the day the goal and outcome are the same—kill patents at all costs regardless of redundancy. Finality can never be achieved by a patent owner. Finality is only achieved when the challenger prevails and the patent is dead.

Allowing multiple identical substantive challenges creates an arbitrage opportunity. A petitioner can file an IPR, absorb a denial, and then pivot to reexamination with no meaningful change in substance. This approach effectively reintroduces the very same problems associated with serial IPR filings—only now through a different procedural door.

Incentivizing Infringers

Compounding the problem is the low threshold required to initiate reexamination. The “substantial new question of patentability” (SNQ) standard is supposed to function as a gatekeeper, but in practice it does not meaningfully screen out weak or repetitive challenges. The petition notes that SNQs are found in over 92% of reexamination requests. That statistic is telling. It means that even when the art is not new, the arguments are not new, and the Office has already declined to institute review in an IPR, the reexamination can still proceed.

The real-world implications extend beyond the USPTO and directly affect parallel litigation. In this case, IBM’s reexamination filing had an immediate and predictable effect: it prolonged the stay of district court proceedings, which is right out of the infringer playbook. What should have been a failed IPR followed by resumed litigation instead became a two-step delay mechanism, which is win-win for alleged infringers who already weaponize procedural delay when they cannot win on substance and merit.

Fighting a war of attrition has been a tried-and-true strategy for alleged infringers for a long time because it has been and continues to be extremely successful. While it is disheartening to see a large patent owner like IBM resort to that tactic themselves, more disheartening is a patent system that enables procedural delay, which universally disadvantages patent owners.

From a strategic standpoint, why would an alleged infringer do anything other than take advantage of this two-step delay strategy? A party can file an IPR with limited downside risk, and apparently no estoppel concerns whatsoever. If the IPR is denied, simply pivot to reexamination and continue to benefit from delay. In this ecosystem, reexamination becomes less about correcting patent validity and more about stringing out litigation as far into the future as possible knowing that at least some percentage of patent owners won’t be able to survive to the end. That is a distortion of purpose, and the system can’t allow procedural gamesmanship to become a substitute for winning on substance.

Art Already Considered by the Examiner and PTAB Cannot Be ‘New’

Even putting aside the failed IPR, in this case the reexamination should fail on its own terms because the prior art is not new. Logically, prior art that is known and simply repackaged from a failed IPR can’t ever possibly meet the threshold of creating a substantially new question of patentability. To create a “substantially new question” doesn’t the prior art need to be new? And if the prior art isn’t new itself doesn’t there need to be something—anything—new about the interpretation or understanding of that prior art? How the same art repackaged into a new procedure suddenly creates reexamination jurisdiction is a real mystery. After all, if something isn’t even new how could it possibly be “substantially new”?

Here the two key references—Dickinson and Hardjono—were already presented to and considered by the examiner during prosecution through information disclosure statements. The third reference, Moulton, is cumulative, adding nothing material beyond what was already before the Office during prosecution. So, the reexamination is not only duplicative of the IPR—it is duplicative of the original examination itself. Duplicative re-litigation of issues and prior art already considered by both the examiner and the PTAB can’t in any honest way support a “new question”, let alone a “substantial new question.”

Sadly, this is not an isolated incident. There is a growing trend of defendants filing copycat reexaminations after failed IPRs. While that tactic is not surprising given increasing IPR denials, the low SNQ threshold that doesn’t actually require there to be a “new question” creates a repeatable and scalable strategy to harass patent owners and delay resolution of patent disputes. If left unchecked, the system risks shifting toward a model where every patent challenge becomes a multi-shot challenge within the Office, where finality is not just elusive but impossible.

USPTO: Stop the Abuse

The uncomfortable reality is that the current framework incentivizes persistence rather than merit. A challenger does not need to prevail on its first attempt; it simply needs to keep the issue alive within the system for long enough to prevail once. Meanwhile, the patent owner must prevail in every battle or they lose the war. How can it make sense to anyone that the patent system in America never reaches a final decision unless and until the patent owner loses? An alleged infringing challenger can lose repeatedly—and on the same arguments and prior art—and still they get the next fresh opportunity to win at which time finality is achieved.

If the PTAB declines to institute an IPR, the Central Reexamination Unit (CRU) should not be allowed to revisit the same challenge under a different procedural label simply because of some fictional notion that an IPR and reexamination are somehow different. Consistency demands that substance control over form. Logic and intellectual honesty demand the USPTO to step in and stop this abuse.

Image Source: Deposit Photos
Image ID: 19011327
Image Author: elly_l