Finding prior art beyond patents and journals and turning it into admissible, persuasive invalidity evidence
1. Introduction
When a company is sued for patent infringement, the most powerful defensive lever is often invalidity. A patent that should never have issued cannot be infringed. Yet issued patents carry a statutory presumption of validity and in district court the challenger must prove invalidity by clear and convincing evidence (35 U.S.C. § 282; Microsoft v. i4i, 2011). That burden makes the quality of prior art decisive. The art the examiner found and considered has already survived scrutiny. The art that wins cases is often the art the examiner never saw.
Conventional invalidity searches look in patent databases and in journals and conference proceedings. They are essential but increasingly insufficient, because plaintiffs and examiners search the same places. Unconventional searches look where the examiner did not: product manuals, archived websites, source code repositories, standards-body contributions, trade-show materials, theses, forum posts, regulatory filings, foreign-language publications and the physical products themselves. This article explains how to plan and execute such searches and how to turn the results into evidence that holds up in district court, the Patent Trial and Appeal Board (PTAB) and the International Trade Commission (ITC).
2. The Legal Backdrop: What Counts as Prior Art and Where
2.1 Statutory categories
For patents governed by the America Invents Act (applications with an effective filing date on or after 16 March 2013), 35 U.S.C. § 102(a)(1) treats as prior art anything patented, described in a printed publication, in public use, on sale or otherwise available to the public before the effective filing date, subject to the one-year grace period exceptions in § 102(b). Section 102(a)(2) covers earlier-filed patent applications by others. For older, pre-AIA patents, the earlier § 102(a), (b) and (g) apply, including prior invention by another.
Section 103 (obviousness) allows combinations of references where the differences between the claimed invention and the prior art would have been obvious to a person of ordinary skill. KSR v. Teleflex (2007) rejected a rigid teaching-suggestion-motivation test and endorsed a flexible approach in which common sense, design incentives and market pressure can supply motivation.
2.2 Forum-specific limits
| Forum | Prior art permitted | Standard of proof | Notes |
| District court | All categories, including public use, on-sale, prior knowledge and system art | Clear and convincing evidence | Jury may weigh credibility of witnesses and physical evidence; § 101 and § 112 defences also available |
| Inter partes review (IPR) | Only patents and printed publications, on §§ 102 and 103 grounds (35 U.S.C. § 311(b)) | Preponderance of the evidence | Petition generally due within one year of service of the complaint (§ 315(b)); discretionary-denial policies have shifted repeatedly, so check current USPTO guidance |
| Post-grant review (PGR) | Any ground of invalidity, including §§ 101 and 112 | Preponderance of the evidence | Available only within nine months of grant, for AIA patents |
| Ex parte reexamination | Patents and printed publications | Substantial new question of patentability | Anonymous filing possible; no estoppel for the requester; patent owner controls the proceeding |
| ITC (Section 337) | Same as district court | Clear and convincing evidence | Fast schedule; discovery timelines compress search work |
Why this matters for search strategy: Because IPR is limited to patents and printed publications, the most unconventional prior art, such as a prior commercial product or public use, can only be deployed in district court or the ITC. A well-designed strategy often uses system art in court while using printed documents that describe the same system in an IPR.
3. Why Unconventional Searches Work
- Examiner constraints. Examiners have limited time and rely mainly on patent classification searching and a fixed set of non-patent literature databases. Technical documentation outside those sources rarely enters the record.
- Fast-moving fields. In software, networking, consumer electronics and biotech, much of the true state of the art lives in product documentation, open-source code, standards drafts and conference talks rather than in patents.
- Priority-date focus. What matters is what was publicly available before the effective filing date. A document hosted on an obscure forum in 2005 can be as invalidating as a patent, if it can be authenticated and shown to be publicly accessible.
- Plaintiff’s blind spots. Patent owners often do not know about earlier products in their own field, including their own prior products placed on sale more than a year before filing.
- Leverage in settlement. A credible, well-documented invalidity case that goes beyond the cited art dramatically shifts negotiation dynamics.
4. A Map of Unconventional Prior Art Sources
4.1 Product manuals, datasheets and marketing collateral
User guides, service manuals, installation guides, datasheets, brochures and release notes can disclose detailed functionality. They are especially valuable when the claimed invention is a feature of a commercial product. Archived copies are available from manufacturers, distributors, libraries, collectors and online marketplaces. Check for publication dates, edition numbers, copyright dates, part numbers and distribution lists.
4.2 Physical products and system art
An actual earlier product (a router, a handset, a medical device, a software release) can be prior art under the ‘public use’ and ‘on sale’ categories. System art requires proof of what the system did before the critical date. Search eBay, surplus dealers, museums, university labs, former employees and customers. Retain the physical item with chain of custody and, where relevant, record its firmware version and operation through documented testing (a ‘working-demonstration’ style exhibit).
4.3 Archived websites and web content
The Internet Archive’s Wayback Machine, web-archiving services and search-engine caches can show what a website disclosed on a given date. Courts generally require authentication; typical approaches include an affidavit from an Internet Archive representative, testimony from a person with knowledge or expert testimony on how the archive operates. Capture all linked resources, as pages often rely on embedded files that the archive may crawl on different dates. Record the dates carefully, because an archive ‘snapshot’ date is not necessarily the date the page was first published.
4.4 Source code and software repositories
Open-source version-control histories (commit logs with timestamps), mailing-list archives, package repositories, release tarballs and bug trackers are valuable in software cases. They show not just that a feature existed but exactly how it was implemented. Be mindful of how timestamps can be altered in rewritten histories; corroborate with independent sources such as mailing-list posts or release announcements.
4.5 Standards bodies and industry consortia
Drafts, contributions, meeting minutes, liaison statements and white papers from organisations such as IEEE, IETF, 3GPP, ETSI, W3C and ISO may qualify as printed publications if they were publicly accessible. IETF RFCs and internet drafts, for instance, are widely archived and dated. Contributions to standards bodies can demonstrate that a feature was known before a patentee’s claimed priority date, even if the patent was filed later and drafted to read on the standard.
4.6 Academic theses, dissertations and course materials
Doctoral and master’s theses held in university libraries can be printed publications if indexed and accessible to the interested public; evidence of cataloguing is often decisive (see In re Cronyn on this point). Lecture slides, posters and course notes may qualify if they were disseminated without restriction. Search institutional repositories and library catalogues.
4.7 Conference presentations, posters and trade-show materials
Public accessibility turns on factors such as the intended audience, whether copies were distributed and whether the material was indexed or catalogued. In re Klopfenstein and later decisions show the importance of evidence on how long a poster was displayed, who attended and whether attendees could copy it. Collect programmes, attendee lists, photographs and witness declarations.
4.8 Foreign-language and regional publications
Patent offices and technical journals in Japan, Korea, China, Germany, Russia and elsewhere contain art that English-language search tools miss. Use machine translation to screen, then obtain certified translations for any reference you will rely on. Regional trade journals and local-language standards publications can be gold mines.
4.9 Regulatory and government documents
FDA submissions and approval documents, clinical trial registries, environmental filings, government procurement documents, defence technical reports (such as DTIC) and public grant abstracts may disclose technical details. Documents obtained under public records laws may be admissible if they were accessible to the public before the critical date.
4.10 Patent applications and non-obvious patent sources
Within patent literature, unconventional means searching unpublished or less-visible categories: foreign priority applications, abandoned applications that became publicly available, utility models, petty patents, defensive publications, Research Disclosure and IP.com prior-art database postings and the file histories of related patents. Also check patents cited in the patentee’s own other patents and the patentee’s earlier product literature.
4.11 Forums, blogs, newsgroups and social media
Usenet archives, developer forums, hobbyist communities, early blogs and maker sites often contain detailed technical descriptions, particularly for software, electronics and consumer technology. Their value depends on authentication and date reliability, so collect corroborating evidence.
4.12 Human sources
Former employees, early customers, academic collaborators, retired engineers and trade-association veterans can lead to documents, products and testimony. Interviews often produce the single most useful reference, such as a customer manual stored in a basement. Witnesses will also be needed to authenticate documents and testify about public use.
5. Designing the Search
- Understand the claims and define the critical date. Identify the effective filing date for each claim, taking into account priority claims and whether the specification supports each claim as of the earlier date. If a claim does not receive the benefit of an earlier application, the critical date for prior art moves later and new art may become available.
- Construe the claims. Draft proposed constructions and search under both the patentee’s broad reading and the narrower reading. The better the match between your art and the broad reading, the more leverage you have, since infringement positions typically require broad constructions.
- Create a feature breakdown. Split each claim into elements, then list synonyms, alternative terminology, historical terms and related technologies. Search in the vocabulary of the earlier era, not today’s.
- Review the file history. Identify art the examiner considered, statements the patentee made to overcome rejections (possible disclaimers) and any claim amendments. Use these to steer the search toward gaps.
- Run conventional searches first. Patent databases and the main non-patent-literature databases establish the baseline and help avoid duplicating the examiner’s work.
- Branch into unconventional sources. Prioritise by field: software and internet (code repositories, archived websites, RFCs); hardware and electronics (manuals, datasheets, physical products, trade shows); life sciences (theses, conference abstracts, regulatory filings, sequence databases); telecommunications (standards drafts and contributions).
- Use interviews and requests. Contact potential witnesses, libraries, archives and companies; send preservation requests; use third-party subpoenas when in litigation.
- Document as you go. Maintain a log of search terms, sources, dates and results, which can support later authentication and demonstrate diligence.
- Evaluate and rank. Chart each reference against each claim element, identify gaps and identify combinations that fill gaps. Consider secondary considerations such as commercial success and copying that a patentee may raise.
6. From Discovery to Admissible Evidence
Finding a promising reference is only half the job. It must be proven to be prior art and that proof has to satisfy rules of evidence.
6.1 Public accessibility and dates
A ‘printed publication’ is one that was disseminated or otherwise made available such that persons interested and ordinarily skilled in the subject matter, exercising reasonable diligence, could locate it (In re Klopfenstein; Blue Calypso v. Groupon, Fed. Cir. 2016). For online materials, show indexing, search accessibility and actual dates. For physical materials, show distribution and catalogue records.
6.2 Authentication and hearsay
Documents must be authenticated (Federal Rule of Evidence 901) and, if offered for their truth, fit a hearsay exception or exclusion. Business records, the ancient-documents exception (under the current federal rule, documents prepared before 1998) and statements offered for the fact of publication rather than the truth are common routes. Plan to use declarations from librarians, archivists or company custodians and obtain certified copies where possible.
6.3 Corroboration of oral testimony
Where invalidity rests on oral testimony about an earlier product or use, the Federal Circuit requires corroboration by independent evidence such as documents, physical exhibits or third-party witnesses (see Woodland Trust v. Flowertree Nursery). Plan for corroboration at the outset by locating invoices, purchase orders, emails, photographs and product samples.
6.4 Enablement of the prior art
To anticipate, a reference must describe the claimed invention and enable a skilled person to practise it. For obviousness, the reference need only be prior art that a skilled person would consider. System art should be supported by evidence of how the system operated, ideally including source code, test records or working units.
6.5 Expert testimony
Invalidity cases typically need a technical expert who can explain the art, map it to the claims and address the level of ordinary skill. Brief the expert early so that the search can be steered by their insights and so that unconventional art is explained in credible technical terms.
7. Tactical Deployment in Litigation and Post-Grant Proceedings
7.1 Timing and disclosure
Many jurisdictions require early service of invalidity contentions, often within months of the complaint (for example, under local patent rules in several US districts). Late discovery of key art can require leave to amend. Start the search before or immediately upon receiving a demand letter or complaint and preserve your ability to file an IPR within the one-year bar.
7.2 Coordinating IPR and district-court strategy
A petitioner in an IPR is estopped, after a final written decision, from asserting in later proceedings any ground it raised or reasonably could have raised in the IPR (35 U.S.C. § 315(e)). Choose IPR grounds carefully and consider holding back system art for district court, while using printed publications that describe the same system in the IPR. Courts also decide whether to stay litigation pending IPR; strong, documented art improves the odds. The current USPTO approach to institution, including discretionary denial and the treatment of parallel litigation, has changed several times in recent years, so check current rules and Director guidance before building a plan around IPR.
7.3 Ex parte reexamination
Reexamination offers a lower-cost alternative and can be filed anonymously. It lacks IPR-style estoppel for the requester, which can be attractive if you want to keep district-court options open, but you also give up the adversarial participation that IPR provides.
7.4 Pairing invalidity with other defences
Invalidity searches complement other defences, including non-infringement, prosecution history estoppel, § 101 ineligibility, indefiniteness, lack of written description or enablement, inequitable conduct, patent exhaustion, licence and, in appropriate cases, prior commercial use under 35 U.S.C. § 273. Information found while searching can support multiple defences. For example, an earlier product may both invalidate and show prior commercial use.
7.5 Using the search in negotiations
A concise claim chart showing that an earlier product or document meets every element, together with a credible plan to prove public accessibility, can lead to early dismissal, low-value settlement or a covenant not to sue. Because plaintiffs weigh litigation cost and risk, demonstrating that you have found art that the Patent Office never considered changes the risk calculation. Avoid overstating the case, since credibility matters in later stages.
8. Practical Tips and Pitfalls
| Pitfall | Why it hurts | Better practice |
| Relying on a single ‘silver bullet’ reference | If it fails on one element or on accessibility, the defence collapses | Build a portfolio: lead reference, backups and combinations |
| Ignoring the critical date | Art that post-dates the effective filing date is useless | Verify date and priority entitlement early |
| Weak authentication of web content | Court excludes the reference | Gather archivist declarations and corroborating sources |
| Uncorroborated witness recollection | Corroboration requirement not met | Collect documents, units and third-party testimony |
| Searching in today’s terminology | Misses historic art | Use period-appropriate vocabulary and classification |
| Overlooking the patentee’s own prior products | Own on-sale and public use may trigger bars | Review the patentee’s product history and marketing |
| Cost blow-out | Unfocused search consumes the budget | Stage the search: baseline, targeted branches, then deep-dive on promising leads |
| Spoliation and preservation failures | Loss of documents and sanctions | Issue litigation holds and preservation requests early |
9. Ethical and Practical Considerations
- Respect confidentiality and privilege in communications with witnesses; consider using a consulting expert for early exploration.
- Do not approach represented parties directly; follow ethics rules on contact with third-party witnesses.
- Be careful with the scope of web scraping and archive access; follow terms of use and legal restrictions.
- Do not alter or reconstruct historical documents; maintain original files, hashes and metadata where possible.
- Remember the duty of candour and Rule 11 obligations when asserting contentions.
10. Case Study Sketch: A Software Feature Patent
A company is sued on a patent claiming a method of synchronising user settings across devices through a cloud profile, with a priority date in 2010. The examiner cited only patents and a handful of journal articles. A broader search turned up three unconventional sources: a product manual for a 2008 consumer device describing profile sync through a vendor server; a snapshot of an open-source project’s wiki and commit history from 2009 showing the same architecture; and an IETF draft from 2009 describing the signalling. The defendant used the manual and IETF draft, both printed publications, in an IPR to challenge the claims on obviousness grounds, while reserving the physical device and source code as system art for the district court. A former engineer of the device maker confirmed the dates and supplied archived firmware. Corroborated by documents and an authenticated archive snapshot, the evidence formed a layered case: if the PTAB did not cancel the claims, the court could still consider the product. The plaintiff, facing parallel risks, settled on favourable terms. (This example is illustrative and not a prediction of outcomes.)
11. Checklist
- Define the effective filing date for each claim; check priority entitlement.
- Construe the claims broadly and narrowly; build a claim-element breakdown.
- Review the file history, cited art and prosecution statements.
- Run baseline patent and non-patent literature searches.
- Search unconventional sources relevant to the field: manuals, products, archives, code, standards, theses, conferences, regulatory records, foreign-language sources.
- Identify and interview human sources; send preservation requests.
- Gather authentication and corroboration for each key reference.
- Prepare claim charts and align the strategy across IPR, reexamination and court.
- Track deadlines: IPR one-year bar, local-rule contentions, PGR nine-month window.
- Retain a technical expert early; maintain a documented search log.
12. Conclusion
Unconventional invalidity searches are one of the most cost-effective ways to change the outcome of patent litigation. They find the art examiners never considered, they exploit the gap between what exists in the real world and what appears in patent databases and they create credible leverage. Success depends on three disciplines: searching creatively in the language and venues of the relevant era; proving public accessibility and dates with admissible, corroborated evidence; and deploying the art in the forum and sequence that best uses its strengths. Teams that plan this early, document carefully and coordinate their legal strategies are best placed to defeat weak claims and to protect their products and customers.
