Prior Art

Prior art is any evidence that an invention — or something that anticipates it or would render it obvious — was already publicly known or otherwise available before the effective filing date (or priority date) of a patent application claiming it. It is the yardstick patent examiners, and later courts or post-grant challengers, use to test whether an invention is genuinely novel and non-obvious, the two substantive bars a claimed invention must clear to be patentable at all. Prior art can take almost any public form: an earlier patent or published patent application, a journal article or preprint, a conference presentation, a thesis, a publicly demonstrated prototype, or a product already on sale. Critically, prior art is not limited to material created by someone else — an inventor’s own public disclosure of their invention, made before they file, can become prior art against their own later patent application.

What Legally Counts as Prior Art

In the United States, prior art is defined by 35 U.S.C. § 102, in its current first-inventor-to-file form under the America Invents Act (AIA), applicable to applications with an effective filing date on or after March 16, 2013:

• 102(a)(1) — public prior art. An invention is not patentable if it was patented, described in a printed publication, in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention. This category explicitly includes disclosures made by the inventor themselves, not only third-party material.
• 102(a)(2) — secret prior art. An invention is not patentable if it was already described in a U.S. patent, a published U.S. patent application, or a published PCT application naming a different inventor, and that earlier application was effectively filed before the claimed invention — even if it had not yet published and so was not actually discoverable by anyone at the time.

For the full timing mechanics behind these two provisions, including the AIA’s effective-date rules and how a public disclosure interacts with the invention-disclosure clock, see 35 U.S.C. § 102: Patent Novelty and Invention Disclosure Timing.

Prior Art and the Novelty vs. Non-Obviousness Tests

Prior art is used two different ways during examination, and the distinction matters for how a reference gets applied:

• Novelty (35 U.S.C. § 102). A single prior-art reference must disclose every element of a claim, arranged as claimed, to defeat novelty — this is called anticipation. If one reference doesn’t say it all, novelty survives even if the reference is very close.
• Non-obviousness (35 U.S.C. § 103). Even where no single reference anticipates the claim, an examiner (or a challenger in litigation or an inter partes review) can combine two or more prior-art references and argue the combination would have been obvious to a person having ordinary skill in the art (a “POSITA”) at the time of filing. Obviousness rejections are where most substantive patent-prosecution disputes happen, because the outcome turns on how a hypothetical skilled person would have read and combined the cited references.

A granted patent isn’t the end of the story either: prior art not found (or not fully considered) during examination is the standard basis for challenging an issued patent’s validity after the fact, including in inter partes review before the USPTO’s Patent Trial and Appeal Board.

Why Prior Art Timing Matters for Researchers and TTOs — Before You Publish

Because an inventor’s own public disclosure can become prior art against their own application, the single highest-stakes decision in academic technology transfer is often not what to patent, but when to file relative to when you plan to publish, present, or otherwise disclose.

U.S. law gives a partial safety net: under 35 U.S.C. § 102(b)(1) and (b)(2), a disclosure made one year or less before the effective filing date, and traceable to the inventor, is excepted from counting as prior art against that inventor’s own U.S. application — commonly called the one-year grace period. Miss that window (as in the first worked example above) and the inventor’s own paper, poster, or talk becomes prior art against their own filing.

The grace period is a U.S.-specific safety net, not a global one. Most major foreign patent offices — including the EPO and China’s CNIPA — apply an absolute novelty standard, with no general grace period for the inventor’s own prior disclosure. A conference talk or journal article that falls comfortably inside the U.S. one-year window can already have destroyed patent rights in Europe or China the moment it became public, because those offices start the novelty clock immediately and don’t forgive inventor-originated disclosures the way U.S. law does. (The EPO’s own exception is narrow: it applies only to evident abuse of the applicant’s rights, or disclosure at specific recognized international exhibitions, and it’s time-limited to six months.) For research institutions that expect to seek patent protection outside the U.S. — which most technology transfer offices at least keep open as an option via a PCT application — the practical rule is to file (typically a provisional patent application) before any public disclosure, not to rely on the U.S. grace period as a fallback.

“Public disclosure” also reaches further than most researchers assume. Helsinn Healthcare S.A. v. Teva Pharmaceuticals USA, Inc., 586 U.S. 123 (2019), confirmed that a commercial sale to a third party who is contractually bound to keep the invention confidential can still trigger the AIA’s on-sale bar under 102(a)(1) — confidentiality of the transaction does not remove it from prior-art status. Purely internal lab discussion, by contrast, is not prior art (see the counter-example above), which is exactly why the line between “discussed only inside the lab” and “disclosed to any outside party, even confidentially” is the one that matters operationally.

This timing pressure also interacts directly with a university’s own internal Bayh-Dole clock. Under 37 CFR § 401.14(c)(2), an institution must elect in writing whether to retain title to a federally funded invention within two years of disclosing it to the funding agency — but where a public disclosure or sale has already started the 35 U.S.C. § 102(b) one-year statutory bar running, the funding agency may shorten that two-year election window to as little as 60 days before the statutory bar expires. In practice, an inventor’s planned conference presentation or paper submission can compress both the patent-filing deadline and the institution’s internal election deadline at the same time — one more reason technology transfer offices ask researchers to submit an invention disclosure (commonly reported to the funding agency via iEdison) well before, not after, a planned publication or presentation date.

Where Prior Art Comes From, and How It Surfaces

Prior art searches (by an inventor’s patent attorney before filing, by a USPTO examiner during prosecution, or by a third party challenging an issued patent) typically draw on:

• Patent databases — USPTO Patent Public Search, Google Patents, Espacenet, and equivalent tools at other national patent offices
• Peer-reviewed journal articles, preprints, and conference proceedings
• Theses and dissertations, which count as “printed publications” once publicly accessible
• Products, devices, or processes already in public use or on sale, including at trade shows and demonstrations
• Grant abstracts and other publicly posted funding-agency records, once publicly available

During examination, a USPTO examiner’s prior-art search results appear in the office actions issued during patent prosecution; an applicant can narrow or amend claims in response to distinguish them from the cited references, or argue the references don’t actually anticipate or render the claims obvious.

Prior Art vs. Related Concepts

Prior art is easy to conflate with a few adjacent ideas it’s worth distinguishing cleanly:

• Prior art vs. a trade secret. A trade secret is information an owner deliberately keeps confidential precisely because publicizing it would destroy its value and, often, its patent-eligibility posture. Choosing trade-secret protection over patenting is sometimes a deliberate strategy specifically to avoid ever creating a prior-art disclosure — the tradeoff being no fixed term of protection, but also no exclusive right if someone else independently discovers or reverse-engineers the same information.
• Prior art vs. patent invalidity. Prior art is the evidence; invalidity is the legal conclusion a court, the PTAB, or an examiner reaches after applying that evidence against a specific claim under 35 U.S.C. §§ 102/103. Not every piece of prior art invalidates a patent — it has to actually anticipate a claim or combine with other references to render it obvious.
• Prior art vs. the cost/timeline decisions around filing. Once a disclosure risk is identified, the practical next step for most academic inventions is usually a provisional patent application (see also a worked example and the real cost breakdown), which secures a filing date cheaply while a full non-provisional application, and any foreign filings, are prepared within the following 12 months. Whether the resulting patent rights are later licensed exclusively, non-exclusively, or to a startup is a separate question, covered in patent licensing.

For the broader technology transfer landscape this term sits within, see the tech transfer pillar page.