Prior Art: What It Is and How It's Used in Patent Disputes

You've been accused of patent infringement. The asserted patent seems broad, covering technology that feels like it's been around for years. Your first thought: "Surely someone did this before the patent was filed."

That "someone did this before" is prior art—and it's one of your most powerful tools for challenging patent validity.

In this article, we explain what prior art is, what types can invalidate patents, where to find it, and how it defeats patent claims through anticipation and obviousness. We specialise in finding prior art for software and AI patents, where much of the relevant material exists outside traditional patent databases—in academic papers on ArXiv, open source code on GitHub, and technical conference proceedings that patent examiners may have missed.

What Is Prior Art?

Prior art is anything that was publicly available before a patent's effective filing date that shows the claimed invention was already known.

The principle is consistent across major jurisdictions. Under 35 U.S.C. § 102 (as amended by the America Invents Act), prior art includes material that was "patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention." The UK Patents Act 1977, Section 2, provides that an invention is new "if it does not form part of the state of the art"—defined as everything made available to the public anywhere in the world before the priority date. Article 54 of the European Patent Convention contains similar novelty requirements.

The purpose is straightforward: patents should only be granted for truly novel inventions. If something was already public knowledge before you filed your patent application, you cannot claim exclusive rights to it. As the USPTO explains, the printed publication provision "prevents an inventor from removing from the public domain that which is already in the possession of the public."

The word "prior" is critical. The prior art must have been publicly available before the patent's priority date—typically the earliest filing date. Material published the day after the priority date, even if years before the patent actually issued, isn't prior art for purposes of challenging that patent.

UK vs US: Key Differences in Prior Art Rules

While the concept of prior art is universal, important differences exist between jurisdictions that affect search strategy and validity arguments.

Aspect	United Kingdom / EPO	United States
System type	First to file (absolute novelty)	First inventor to file (AIA 2013)
Grace period	None—any public disclosure before filing destroys novelty	One year grace period for inventor's own disclosures
Prior art date	Priority date	Effective filing date
Novelty standard	Strict absolute novelty (UK Patents Act s.2)	Novelty with grace period exception (35 U.S.C. § 102)
Secret prior use	Not prior art unless made available to public	May qualify as prior art under certain conditions
Obviousness term	"Inventive step" (Article 56 EPC)	"Non-obvious" (35 U.S.C. § 103)
Obviousness test	Problem-solution approach (EPO), Windsurfing/Pozzoli (UK)	Graham factors, KSR flexible approach
Published applications	Prior art from filing date for novelty only	Prior art from filing date for novelty and obviousness

The Grace Period Difference

This distinction matters significantly. In the US under the AIA, there is a one year grace period for disclosures made by the inventor or a joint inventor. If you publicly disclose your own invention and file a patent application within one year of that disclosure, your own disclosure typically won't count as prior art against you.

In the UK and Europe, no such grace period exists. Any public disclosure before the filing date—including the inventor's own disclosure—destroys novelty and can invalidate the patent. This "absolute novelty" requirement under Section 2 of the UK Patents Act 1977 means that researchers and companies seeking UK or European patent protection must file before any public disclosure, including at academic conferences, in journals, or in press releases.

We regularly see patent validity challenges succeed in Europe based on an inventor's own pre-filing publication that wouldn't be prior art in the US. This is critical for technology companies with international patent portfolios.

Obviousness Standards

The UK and EPO apply the "problem-solution approach" to assess inventive step:

1. Determine the closest prior art
2. Establish the objective technical problem
3. Consider whether the claimed invention would have been obvious to the skilled person

UK courts follow the Windsurfing/Pozzoli test refined in Pozzoli SpA v BDMO SA 2007 EWCA Civ 588:

1. Identify the notional person skilled in the art and their common general knowledge
2. Identify the inventive concept of the claim
3. Identify differences between cited prior art and the inventive concept
4. Decide whether those differences constitute steps that would have been obvious to the skilled person or require any degree of invention

US courts apply the Graham factors from Graham v. John Deere Co. (1966): scope and content of prior art, differences between prior art and claims, level of ordinary skill in the art, and secondary considerations. The Supreme Court's KSR Int'l Co. v. Teleflex Inc. (2007) decision liberalised obviousness analysis, allowing courts to apply "common sense" and recognise that combining known elements with predictable results is typically obvious.

Types of Prior Art

Prior art takes many forms. Understanding what qualifies is essential for finding it and evaluating its strength.

Patents and Published Patent Applications

Previously granted patents and published patent applications from any country are prior art if published before the relevant priority date. This includes patents that were never commercially practised, patents that have since expired or been abandoned, and patent applications that were eventually rejected.

Under both UK and US law, an unpublished pending application can be prior art for novelty purposes from its filing date—even if published after the later patent's priority date. This creates "secret prior art" that applicants cannot discover until publication, typically 18 months after filing.

Patent databases make these relatively easy to find. We search USPTO (United States), Espacenet/EPO (Europe and worldwide patents), WIPO (international applications), and national patent offices (UK IPO, Japan Patent Office, China National Intellectual Property Administration).

Printed Publications

The term "printed publication" is interpreted broadly. For something to qualify as a "printed publication," it must have been "disseminated or otherwise made available to the extent that persons interested and ordinarily skilled in the subject matter or art, exercising reasonable diligence, can locate it." Public accessibility is the touchstone—not whether something was literally printed.

This includes:

Academic papers: Journal articles, conference proceedings, dissertations, technical reports. For AI and machine learning, ArXiv is particularly important—researchers typically publish preprints there with timestamps showing publication dates.

Technical standards: IEEE, W3C, IETF, ISO specifications describe technical implementations in detail with clear publication dates.

Books and textbooks: Technical references and engineering handbooks.

Trade publications: Industry magazines, technical newsletters, product announcements.

Product documentation: User manuals, datasheets, technical white papers, application notes.

The EPO follows similar principles. Under EPO guidelines, a document is publicly available if "at least one member of the public could gain knowledge of the content of the document and there was no obligation to keep the content confidential." The burden is on the party challenging validity to prove public accessibility.

Open Source Code and Software Repositories

For software patents, publicly available source code is prior art. GitHub, GitLab, and SourceForge host code repositories with commit timestamps showing exactly when code was made public.

We regularly find invalidating prior art in open source projects. A software patent claiming a specific algorithm, data structure, or architectural pattern may have prior art in a GitHub repository that implemented the same approach months or years before the patent application was filed. The commit history provides verifiable publication dates.

This is particularly relevant for AI and machine learning patents. Open source ML libraries (TensorFlow, PyTorch, scikit-learn), model implementations on GitHub, and shared code accompanying academic papers create a rich landscape of prior art that patent examiners searching only patent databases will miss.

Web Archives and Online Content

Websites, blog posts, technical documentation, and online publications can be prior art if they were publicly accessible and their publication date can be verified.

The Wayback Machine (Internet Archive) is frequently used to establish publication dates for web content. If a webpage was archived on a specific date before the patent's priority date, that archive can demonstrate the content was publicly available.

However, Wayback Machine evidence alone may be insufficient. Recent Patent Trial and Appeal Board (PTAB) decisions emphasise that you must demonstrate not just that content was technically accessible (archived), but that it was publicly accessible to persons ordinarily skilled in the art. This may require additional evidence showing the webpage was indexed by search engines, linked from other sites, or otherwise discoverable by interested parties. The Federal Circuit in Valve Corporation v. Ironburg Inventions Ltd. (2021) took judicial notice of Wayback Machine captures as evidence of prior art printed publications, but with supporting evidence of accessibility.

Conference Presentations and Academic Preprints

For AI, machine learning, and computer science patents, academic conferences are critical prior art sources.

Major conferences (NeurIPS, ICML, ICLR, CVPR, ICCV for AI and computer vision; ACL, EMNLP for NLP) publish proceedings with timestamps. Papers accepted to these conferences are often available as preprints on ArXiv before the conference, creating even earlier prior art dates.

Workshop papers, poster presentations, and technical talks also qualify as prior art if there's documentation of what was presented and when. Conference programmes, archived presentation slides, and video recordings can establish publication dates.

How Prior Art Defeats Patents

Prior art invalidates patents through two primary legal mechanisms: anticipation (lack of novelty) and obviousness (lack of inventive step).

Anticipation (Lack of Novelty)

A patent claim is anticipated—and therefore invalid for lack of novelty—if a single prior art reference discloses every element of the claim.

In the US, the Federal Circuit has clarified that the prior art must disclose elements "arranged as in the claim," not merely present in different contexts within the reference. Anticipation cannot be proven by combining multiple separate teachings found within a single reference—the reference must show all limitations already combined together as claimed. See Microsoft Corp. v. Biscotti Inc., 878 F.3d 1052 (Fed. Cir. 2017).

In the UK, the test for novelty under Section 2 of the Patents Act 1977 asks whether the prior art discloses the claimed invention such that a skilled person would "at once perceive, understand and be able to practically apply the disclosure" without the need for further experimentation. The disclosure must be "enabling"—it must teach the skilled person how to perform the invention.

Example: Patent claims "method for image classification using convolutional neural network with attention mechanism to weight feature importance." A single prior art paper describes the same method—CNN with attention for weighting features in image classification—published before the priority date. Every element is disclosed and arranged as claimed. The claim is anticipated.

The prior art doesn't require identical terminology. It must describe the same technical concept such that a person having ordinary skill in the art would recognise it teaches every claim element.

Obviousness (Lack of Inventive Step)

Even if no single reference anticipates a claim, it can still be invalid if the invention would have been obvious to a person having ordinary skill in the art based on combining prior art references.

In the US, the key question is whether it would have been obvious to combine prior art references. Factors include:

• Whether there was a reason or motivation to combine (not necessarily explicit—can be based on common sense, ordinary innovation, or recognition that combining solves a known problem)
• Whether combining would require only ordinary skill
• Whether results would be predictable
• Whether prior art teaches away from the combination

Example: Patent claims method combining techniques A, B, and C. One prior art reference (academic paper from 2 years before priority date) describes A+B. Another reference (conference paper from 1 year before) describes C in the same technical field. If combining them would have been obvious to someone skilled in the art with predictable results, the claim is invalid as obvious—even though no single reference showed all three.

In the UK and at the EPO, the "inventive step" analysis under the problem-solution approach asks:

1. What objective technical problem does the invention solve starting from the closest prior art?
2. Would the claimed solution have been obvious to the skilled person trying to solve that problem?

UK courts have noted that something may be "obvious to try" with a reasonable expectation of success, which supports an obviousness finding. However, mere workshop modifications or routine experimentation expected of skilled practitioners do not confer inventive step.

Where to Search for Prior Art

Finding strong prior art requires searching comprehensively across multiple databases and publication types.

Patent Databases

USPTO Patent Public Search: The USPTO's modern web-based search tool provides access to US patents and applications dating back to 1790, with both basic keyword search and advanced query interfaces.

Espacenet (EPO): The European Patent Office provides free access to over 160 million patent documents worldwide, including European patents, WIPO international applications, and patents from participating national offices.

Google Patents: Aggregates patent data from USPTO, EPO, and WIPO. Google's Prior Art Finder tool automatically identifies key phrases from patent text and searches simultaneously across Google Patents, Google Scholar, Google Books, and the broader web to locate relevant prior art—a useful starting point.

UK Intellectual Property Office: The UK IPO provides free searchable access to UK patents, with advanced search tools for patent, trade mark, and design searches.

National patent offices: Japan Patent Office (JPO), China National Intellectual Property Administration (CNIPA), Korean Intellectual Property Office (KIPO) provide access to their respective patent databases.

Academic and Technical Publications

ArXiv: The essential resource for AI, machine learning, physics, mathematics, and computer science prior art. Researchers publish preprints with precise timestamps. We find invalidating prior art on ArXiv regularly for AI and ML patents—academic publications that predate patent applications and describe the same technical approaches.

Google Scholar: Searches academic papers, articles, theses, books, and conference proceedings across all disciplines.

IEEE Xplore: Technical papers in electrical engineering, computer science, electronics, and telecommunications. IEEE conference proceedings are major prior art sources for software and hardware patents.

ACM Digital Library: Computer science, software engineering, and information technology research. ACM conference papers (SIGCHI, SIGGRAPH, etc.) describe software innovations that may predate patents.

University repositories: Theses, dissertations, and technical reports from university libraries, often publicly accessible online.

British Library EThOS: The UK's national thesis service providing access to over 500,000 doctoral theses.

Technical Standards

IEEE, W3C, IETF, ISO, and ETSI publish technical standards that serve as detailed prior art. These standards documents describe technical implementations with precision and have clear publication dates. Standards-essential patents must be evaluated against standard specifications that may themselves constitute prior art.

Open Source Repositories and Package Managers

GitHub, GitLab, SourceForge: Public code repositories with commit timestamps showing when code was published. For software patents, dated implementations are definitive prior art.

Package repositories: npm (JavaScript), PyPI (Python), Maven (Java), CRAN (R), RubyGems—all maintain versioned packages with publication dates and source code.

Why Software and AI Prior Art Requires Specialised Searching

For AI and machine learning patents, the prior art landscape differs fundamentally from traditional technology areas.

Academic researchers publish on ArXiv, present at major conferences (NeurIPS, ICML, CVPR, ICLR), and share implementations on GitHub. They don't typically file patents before publication. This means:

• A software patent claiming a machine learning technique may have been fully described in a research paper published on ArXiv six months before the patent application
• An algorithm covered by a patent might be implemented in an open source library with GitHub commits predating the priority date
• Conference papers and technical blog posts from research labs may have disclosed the invention publicly

Finding this prior art requires domain expertise and knowing:

• Where AI researchers publish their work
• What terminology researchers use (often different from patent language)
• How to evaluate whether an academic paper or code implementation actually discloses the claimed invention
• Which conferences and journals are reputable and establish public accessibility

We conduct software and AI prior art searches by examining not just patent databases but the full technical landscape: ArXiv preprints, major conference proceedings, open source repositories, technical blogs from research labs (Google Research, Meta AI, Microsoft Research, DeepMind), and academic databases.

The Technical Expert's Role in Prior Art Searching

Domain Knowledge and Comprehensive Searching

Finding strong prior art requires knowing where to look in your specific technology domain. Generic patent search firms search patent databases. Technical experts with domain knowledge search where innovations in that field are actually published.

For AI patents: ArXiv, NeurIPS/ICML/CVPR conference papers, GitHub repositories, ML research lab blogs.

For telecommunications: IEEE standards, 3GPP specifications, ETSI documents, conference proceedings, technical working group documents.

For web technologies: W3C specifications, browser implementation code repositories, IETF RFCs.

We apply the USPTO's recommended multi-step search methodology, starting with brainstorming technical terms, conducting keyword searches, using Cooperative Patent Classification (CPC) codes for focused searches, and examining both patent and non-patent literature comprehensively.

Technical Evaluation and Element Mapping

Not all prior art that appears relevant actually discloses the claimed invention. Technical evaluation determines whether a reference truly anticipates (single reference shows all elements arranged as claimed) or whether combining references would be obvious.

We create claim charts mapping prior art to patent claims element by element, showing how prior art teaches each claimed limitation. This is the same element by element analysis required for infringement claim charts, but applied to prior art references instead of accused products.

Obviousness Analysis

Determining whether combining prior art references would be obvious requires technical judgement informed by KSR factors (US) or the problem-solution approach (UK/EPO):

• Was there motivation to combine? (Not necessarily explicit—can be based on common sense, ordinary innovation, or recognition that combining solves a known problem)
• Would combining require more than ordinary skill?
• Would results be predictable?
• Does prior art teach away from the combination?

This requires understanding how practitioners in the field approach problems and what would have been routine versus inventive at the relevant time.

Costs and Practical Realities

What Prior Art Searching Costs

Service Level	UK Cost Range	US Cost Range	Typical Deliverable
Basic patentability search	£1,500–£3,500	$2,000–$5,000	Patent database search with summary report
Comprehensive validity search	£5,000–£15,000	$8,000–$25,000	Patent and non-patent literature search with claim charts
Specialised AI/software search	£8,000–£20,000	$10,000–$30,000	Includes ArXiv, GitHub, conference proceedings analysis
Full invalidity opinion	£15,000–£40,000	$25,000–$75,000	Search plus legal opinion from patent counsel

Costs vary significantly based on technology complexity, number of claims analysed, urgency, and whether litigation is involved. Litigation-quality prior art searching that can withstand discovery challenges and expert testimony costs substantially more than preliminary patentability searches.

Timeline Expectations

Stage	Typical Duration
Initial search scoping	1–2 days
Comprehensive prior art search	2–4 weeks
Technical analysis and claim charting	1–3 weeks
Expert report preparation	2–4 weeks
IPR petition prior art preparation	4–8 weeks

For Inter Partes Review (IPR) petitions in the US, the one year deadline from service of an infringement complaint creates real time pressure. Prior art searching should begin immediately upon receiving an infringement allegation.

When Prior Art Searching Is Worth the Investment

Prior art searching typically makes sense when:

• You're accused of infringement — Invalidity is a complete defence. Finding strong prior art can invalidate the asserted patent entirely or force favourable settlement
• You're considering a licence — Understanding patent validity informs whether licensing makes sense and at what terms
• You're evaluating an acquisition — Patent portfolio due diligence should include validity assessment of key patents
• You're filing a patent application — Understanding existing prior art helps craft claims that are both valid and commercially valuable
• You're considering an IPR petition — IPR success rates are higher with strong prior art; professional searching improves odds significantly

In our experience, comprehensive prior art searching pays for itself when it identifies invalidity arguments that change litigation or licensing outcomes. A £15,000 search that defeats a patent asserter's £500,000 damages claim is obviously worthwhile. The challenge is that outcomes aren't guaranteed—but the probability of finding useful prior art is substantially higher with thorough, expert searching.

Critical Mistakes to Avoid

What Not to Do When Searching for Prior Art

Mistake	Why It's Problematic	Better Approach
Searching only patent databases	Misses academic papers, open source code, standards documents	Search non-patent literature comprehensively
Stopping at the first relevant reference	Missing stronger or more on-point prior art	Continue searching even after finding seemingly good references
Ignoring non-English prior art	Foreign language publications are valid prior art	Search Chinese, Japanese, German, French patent databases
Relying solely on keyword searches	Inventors use different terminology than researchers	Use CPC classification codes and technical synonyms
Assuming Wayback Machine timestamps are sufficient	PTAB may require additional evidence of public accessibility	Gather supporting evidence (search engine indexing, links, citations)
Searching after claim construction	Missing prior art because of narrow interpretation	Begin searching early; iterate as claim interpretation develops
Using prior art without technical evaluation	Reference may not actually teach all claim elements	Map prior art to claims element by element before relying on it

Common Prior Art Quality Problems

Ambiguous dates: Prior art with unclear publication dates is vulnerable to challenge. Ensure you can prove when material was publicly available.

Accessibility questions: Material must have been accessible to persons skilled in the art. Internal documents, confidential research, or materials under NDA don't qualify.

Missing elements: A reference that describes 90% of a claim is not anticipating prior art. Every element must be present for anticipation.

Hindsight bias: Don't read the patent first and then interpret prior art through that lens. Obviousness must be evaluated from the perspective of the skilled person before seeing the patent.

Decision Framework: When and How to Use Prior Art

Should You Conduct a Prior Art Search?

Yes, prioritise searching when:

• You've been accused of patent infringement (invalidity is a complete defence)
• The potential damages or licensing costs exceed £50,000
• You're considering an IPR petition (one year deadline from complaint)
• Key patents block your product roadmap
• You're acquiring a company and patents are material to value

Maybe, depending on circumstances:

• You're evaluating a licence offer (weigh search cost against licence cost)
• You're filing a patent application (prior art informs claim strategy)
• You're conducting freedom to operate analysis (validity affects risk assessment)

Probably not:

• The patent is already expired
• You can easily design around the patent
• Licensing costs are minimal
• You have stronger defences available (non-infringement, exhaustion)

Validity Challenge Venue Comparison

Venue	Cost Range	Timeline	Standard	Success Rate
IPR (US PTAB)	$300,000–$600,000	12–18 months	Preponderance (more likely than not)	~65% for instituted petitions
US District Court	$2–5 million (litigation)	2–4 years	Clear and convincing evidence	~30% overall
UK Patents Court	£500,000–£2 million	12–24 months	Balance of probabilities	Varies
UK IPEC	£50,000–£100,000 (capped)	6–12 months	Balance of probabilities	Varies
EPO Opposition	€50,000–€150,000	2–4 years	Balance of probabilities	~30% revocation rate

The standard of proof matters significantly. In US district court litigation, patents are presumed valid and the challenger must prove invalidity by "clear and convincing evidence." At the PTAB in IPR proceedings, the lower "preponderance of the evidence" standard makes invalidation easier—one reason IPR has become the preferred venue for validity challenges in the US.

Quality of Prior Art

Strong Prior Art Characteristics

Verifiable publication date: The prior art must have a provable publication date before the patent's priority date. This might be a journal publication date, ArXiv timestamp, GitHub commit date, or Wayback Machine archive showing web accessibility on a specific date.

Clear technical description: The prior art explicitly describes the claimed invention or its elements using technical language. A person skilled in the art would recognise it teaches what the patent claims.

Credible, publicly accessible source: Peer-reviewed journals, major conference proceedings, granted patents, published technical standards, and established research institutions carry more weight than obscure blogs or unverifiable sources. The source must demonstrate public accessibility to persons ordinarily skilled in the art.

Sufficient detail: For anticipation, the reference must enable someone skilled in the art to practise the invention. Vague descriptions without implementation details are weaker.

Weak Prior Art Characteristics

Ambiguous or unverifiable dates: If you cannot prove when something was published, it's difficult to use as prior art. Wayback Machine archives alone may be insufficient without additional evidence of public accessibility and discoverability.

Vague descriptions: Prior art that generally discusses a problem but doesn't describe the specific claimed solution. Or prior art that mentions concepts without technical implementation details.

Requires excessive inference: If proving anticipation requires assuming the prior art must have used a certain approach but doesn't explicitly describe it, the prior art is vulnerable to challenge.

Questionable public accessibility: Material that may have existed but wasn't disseminated to persons interested and skilled in the art. Internal corporate documents, confidential research, or materials under NDA don't qualify.

Conclusion

Prior art is a strategic tool for invalidating patents, conducting due diligence, and informing patent filing strategies. Finding strong prior art that actually defeats patent claims requires comprehensive searching across patent and non-patent literature and technical evaluation of whether prior art teaches claimed inventions.

For software and AI patents especially, the most relevant prior art often exists in academic publications, open source repositories, and technical standards rather than patent databases. We conduct comprehensive prior art searches examining ArXiv preprints, GitHub repositories, major conference proceedings (NeurIPS, ICML, CVPR, IEEE conferences), technical standards, and research lab publications.

We have the technical expertise to evaluate whether prior art actually discloses claimed inventions and to map prior art to claims element by element for anticipation or obviousness arguments.

If you're facing patent validity challenges or need comprehensive prior art searching, technical experts who understand your technology domain can identify invalidity arguments that change case outcomes.

This is general information about prior art, not legal advice for specific patent disputes. Work with qualified patent counsel and consider engaging technical experts with domain expertise for comprehensive prior art analysis.

---

Sources

Primary Legislation and Rules

1. 35 U.S.C. § 102 — Conditions for patentability; novelty. Defines prior art as material "patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date." https://www.law.cornell.edu/uscode/text/35/102

2. 35 U.S.C. § 103 — Conditions for patentability; nonobvious subject matter. Sets out the obviousness standard requiring analysis of whether claimed invention would have been obvious to one of ordinary skill. https://www.law.cornell.edu/uscode/text/35/103

3. AIA 35 U.S.C. § 102(b)(1)(A) — Statutory grace period exception for inventor-originated disclosures made one year or less before effective filing date.

4. UK Patents Act 1977, Section 2 — Novelty. "An invention shall be taken to be new if it does not form part of the state of the art." https://www.legislation.gov.uk/ukpga/1977/37/section/2

5. UK Patents Act 1977, Section 3 — Inventive step. "An invention shall be taken to involve an inventive step if it is not obvious to a person skilled in the art." https://www.legislation.gov.uk/ukpga/1977/37/section/3

6. European Patent Convention, Article 54 — Novelty. Establishes novelty requirement and defines state of the art for European patents. https://www.epo.org/en/legal/epc/2020/a54.html

7. European Patent Convention, Article 56 — Inventive step. Sets out inventive step standard for EPO examination. https://www.epo.org/en/legal/epc/2020/a56.html

USPTO Guidance

8. USPTO MPEP § 2128 — "Printed Publications" as Prior Art. Establishes that public accessibility is the touchstone for determining whether a reference constitutes a printed publication, with burden on patent challenger to prove both existence and public accessibility. https://www.uspto.gov/web/offices/pac/mpep/s2128.html

9. USPTO MPEP § 2131 — Anticipation — Application of 35 U.S.C. 102. Provides guidance on anticipation analysis requirements. https://www.uspto.gov/web/offices/pac/mpep/s2131.html

10. USPTO MPEP § 2141 — Examination Guidelines for Determining Obviousness Under 35 U.S.C. 103. Details post-KSR obviousness analysis framework. https://www.uspto.gov/web/offices/pac/mpep/s2141.html

11. USPTO MPEP § 2153.01 — Prior Art Exception Under AIA 35 U.S.C. 102(b)(1)(A) (Grace Period Inventor-Originated Disclosure Exception). Explains one-year grace period for inventor's own disclosures. https://www.bitlaw.com/source/mpep/2153-01.html

12. USPTO, "Multi-Step U.S. Patent Search Strategy" — Official guidance recommending 7-step methodology: brainstorm terms, keyword searching, classification searching, review and iteration. https://www.uspto.gov/patents/search/patent-search-strategy

EPO Guidelines

13. EPO Guidelines for Examination, Part G, Chapter IV — State of the Art. Defines what constitutes prior art under EPO practice and establishes public availability standards. https://www.epo.org/en/legal/guidelines-epc/2024/g_iv.html

14. EPO Guidelines for Examination, Part G, Chapter VII — Inventive Step. Describes problem-solution approach to assessing inventive step. https://www.epo.org/en/legal/guidelines-epc/2024/g_vii.html

15. EPO Guidelines for Examination, Part G, Chapter VII, Section 5 — Problem-Solution Approach. Detailed guidance on applying the three-step problem-solution analysis. https://www.epo.org/en/legal/guidelines-epc/2024/g_vii_5.html

UK IPO Guidance

16. UK IPO Manual of Patent Practice, Section 2 — Novelty. Official UK guidance on novelty assessment and state of the art. https://www.gov.uk/guidance/manual-of-patent-practice-mopp/section-2-novelty

17. UK IPO Manual of Patent Practice, Section 3 — Inventive Step. UK guidance on inventive step assessment applying Pozzoli framework. https://www.gov.uk/guidance/manual-of-patent-practice-mopp/section-3-inventive-step

Supreme Court and Federal Circuit Case Law (US)

18. KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) — Supreme Court decision rejecting rigid teaching-suggestion-motivation test for obviousness and allowing more flexible common sense analysis. https://www.law.cornell.edu/supct/html/04-1350.ZO.html

19. Graham v. John Deere Co., 383 U.S. 1 (1966) — Supreme Court decision establishing framework for obviousness analysis (scope of prior art, differences from claims, level of ordinary skill, secondary considerations). https://supreme.justia.com/cases/federal/us/383/1/

20. Microsoft Corp. v. Biscotti Inc., 878 F.3d 1052 (Fed. Cir. 2017) — Federal Circuit clarification that for anticipation, prior art must disclose all elements "arranged as in the claim," not just present in different contexts within reference. https://casetext.com/case/microsoft-corp-v-biscotti-inc-1

21. Valve Corporation v. Ironburg Inventions Ltd., 8 F.4th 1364 (Fed. Cir. 2021) — Federal Circuit took judicial notice of Wayback Machine captures as evidence of prior-art printed publications with supporting accessibility evidence.

22. In re Winslow, 365 F.2d 1017 (CCPA 1966) — Established that obviousness must be assessed from the perspective of one skilled in the art aware of all pertinent prior art.

23. In re GPAC Inc., 57 F.3d 1573 (Fed. Cir. 1995) — Established that a document is publicly accessible if it was disseminated or made available to extent that persons skilled in the art could locate it.

UK Court Decisions

24. Pozzoli SpA v BDMO SA 2007 EWCA Civ 588 — Court of Appeal refined Windsurfing test for inventive step into four-step Pozzoli framework applied in UK patent validity cases.

25. Windsurfing International Inc v Tabur Marine (Great Britain) Ltd 1985 RPC 59 — Established original four-step test for inventive step in UK patent law.

26. Synthon BV v SmithKline Beecham plc 2005 UKHL 59 — House of Lords clarified requirements for anticipation: prior disclosure must be enabling (skilled person could perform the invention).

27. Generics (UK) Ltd v H Lundbeck A/S 2009 UKHL 12 — House of Lords guidance on obviousness and "obvious to try" standard.

PTAB and IPR Guidance

28. USPTO, "Inter Partes Review" — Overview of IPR procedure, standards, and timelines. https://www.uspto.gov/patents/ptab/inter-partes-review

29. USPTO PTAB Trial Statistics (FY2024) — Institution and final written decision statistics showing approximately 65% invalidation rate for instituted petitions.

30. PTAB Litigation Blog, "Proceed With Caution When Using Wayback Machine Prior Art" (2024) — Analysis of PTAB decisions requiring evidence beyond Wayback Machine archives to prove public accessibility, citing First Solar Inc. v. Rovshan Sade (2023). https://www.ptablitigationblog.com/proceed-with-caution-when-using-wayback-machine-prior-art/

Patent Database Resources

31. EPO Espacenet — Free access to over 160 million patent documents worldwide. https://worldwide.espacenet.com/

32. Google Patents About — Overview of Google Patents coverage and Prior Art Finder tool that automatically searches multiple sources. https://support.google.com/faqs/answer/2539193

33. USPTO Patent Public Search — Official USPTO search tool with full text searching of US patents and applications. https://ppubs.uspto.gov/pubwebapp/

34. UK IPO Patent Search — UK Intellectual Property Office patent database. https://www.gov.uk/search-for-patent

35. WIPO PATENTSCOPE — World Intellectual Property Organization international patent search. https://patentscope.wipo.int/

Academic and Technical Publication Resources

36. ArXiv — Open access preprint server for physics, mathematics, computer science, AI/ML research with verifiable publication timestamps. https://arxiv.org/

37. IEEE Xplore — Technical literature in electrical engineering, computer science, and electronics. https://ieeexplore.ieee.org/

38. ACM Digital Library — Computer science and information technology research publications. https://dl.acm.org/

39. Google Scholar — Cross-disciplinary academic search covering papers, theses, books, and conference proceedings. https://scholar.google.com/

40. British Library EThOS — UK thesis repository with access to over 500,000 doctoral theses. https://ethos.bl.uk/

Standards Organisations

41. W3C Technical Reports — Web standards specifications with publication dates. https://www.w3.org/TR/

42. IETF RFCs — Internet Engineering Task Force specifications for internet protocols. https://www.ietf.org/standards/rfcs/

43. IEEE Standards — Electrical and electronics engineering standards. https://standards.ieee.org/

44. ETSI Standards — European Telecommunications Standards Institute publications. https://www.etsi.org/standards

Cost and Practice Information

45. AIPLA 2023 Report of the Economic Survey — Industry data on patent litigation and prosecution costs in the United States. American Intellectual Property Law Association annual survey.

46. UK IPO, "Resolving patent disputes" — Guidance on UK patent dispute resolution options and costs. https://www.gov.uk/guidance/resolving-patent-disputes

47. CIPA Guide to the Patents Acts (9th edition) — Authoritative UK patent law treatise covering validity challenges and prior art standards.

48. Chisum on Patents — Leading US patent law treatise covering anticipation, obviousness, and prior art standards (Matthew Bender).