What is inter partes review, and how is it different from challenging a patent in court? Inter partes review (IPR) is an administrative trial conducted by the Patent Trial and Appeal Board (PTAB) inside the USPTO, created by the America Invents Act. In an IPR, a party that does not own the patent files a petition asking the Board to cancel one or more of the patent's claims as unpatentable. The proceeding is deliberately narrow: it can only be used to argue that claims are invalid for lack of novelty or for obviousness, and only on the basis of prior-art patents and printed publications — not on other validity grounds and not using other categories of evidence. That narrowness is what distinguishes it from a district-court validity challenge, where a broader range of defenses is available.

The authorization comes from 35 U.S.C. 311, which establishes who may file and on what basis. The statute provides:

a person who is not the owner of a patent may file with the Office a petition to institute an inter partes review of the patent. The Director shall establish, by regulation, fees to be paid by the person requesting the review, in such amounts as the Director determines to be reasonable, considering the aggregate costs of the review.— 35 U.S.C. § 311(a), source

Two limits in the rest of section 311 define the proceeding's scope. Subsection 311(b) restricts the grounds: a petitioner may request to cancel claims "only on a ground that could be raised under section 102 or 103" — that is, novelty or obviousness — "and only on the basis of prior art consisting of patents or printed publications." So an IPR cannot be used to argue that claims are indefinite, lack written description, or fail the eligibility test; those are not 102/103 grounds. Nor can it rest on prior public use or on-sale activity, because those are not patents or printed publications. Subsection 311(c) sets timing, generally barring a petition until after a window following the patent's issuance or any post-grant review. The combined effect is a focused, document-based validity challenge.

Why IPR matters for technology-dense patents

For sectors where patents are numerous and densely cited — semiconductors among them — IPR is a significant feature of the IP risk landscape because it offers an administrative route to test an asserted patent's claims against the published prior-art record. A company facing assertion of a semiconductor patent can petition for IPR to challenge the claims on the documentary prior art, and the proceeding runs before technically trained administrative patent judges under statutory deadlines. Because the grounds are confined to 102/103 and the evidence to patents and publications, IPR is best suited to claims whose closest prior art is itself documented in earlier patents or technical publications — a common situation in a field with a deep patent literature. The proceeding's outcome is a written decision on whether the challenged claims are unpatentable, which can cancel claims that survive only on the strength of the issued presumption.

The narrowness is the design, not a limitation to lament

It is worth dwelling on why section 311 confines IPR so tightly, because the confinement is deliberate and shapes how the proceeding is used. By restricting grounds to sections 102 and 103 and evidence to patents and printed publications, the statute keeps IPR to the kind of validity question that can be resolved on a documentary record without live fact disputes about, say, what was secretly used or offered for sale. Public-use and on-sale challenges, which depend on testimony about events, are left to forums equipped to weigh that evidence; written-description, enablement, and eligibility challenges, which turn on the specification and the law of eligibility rather than on prior-art documents, are likewise outside IPR's lane. The result is a proceeding built for speed and technical focus: the contest is whether earlier patents and publications anticipate or render obvious the challenged claims, decided by judges who read the art. This design is why IPR became a heavily used tool in technically dense fields — the validity attack most likely to succeed against a densely cited patent is often exactly the documentary 102/103 attack that IPR is built to adjudicate. Understanding the statute's deliberate narrowness is understanding why a challenger chooses IPR for some patents and court for others.

How it fits with court litigation

IPR runs in parallel to, and interacts with, district-court patent litigation. A defendant sued on a patent may petition for IPR on the same patent, and the existence of the statutory presumption of validity in court (35 U.S.C. 282) is one reason a challenger may prefer the administrative forum's focused prior-art test. The two forums apply the proceeding's own standards and reach their own results; an IPR can cancel claims, narrow what remains assertable, or leave the claims standing. The statute confines IPR to its lane — non-owner petitioner, 102/103 grounds, patents-and-publications prior art — and leaves the broader set of validity and infringement questions to the appropriate forum. A real semiconductor patent such as United States Patent US10672707B2, "Low aspect ratio interconnect," would, if challenged in an IPR, be tested only against earlier patents and printed publications on the novelty and obviousness of its claims — not against other validity theories.

What the record shows is the statute's design. Under 35 U.S.C. 311, a non-owner may petition the USPTO to institute inter partes review of an issued patent, the Director sets the fees, and the challenge is confined by subsection 311(b) to grounds under sections 102 and 103 and to prior art consisting of patents or printed publications. IPR is therefore a focused, document-based administrative test of patent validity, distinct in scope from a district-court challenge. The statute does not predict any outcome — whether a given patent's claims survive an IPR depends on the prior art and the Board's decision — but it fixes who may bring the challenge and on what narrow grounds, which is the first thing an IP reader needs to know when assessing how an issued semiconductor patent can be contested.