In-depth analysis of USPTO and EPO developments, prosecution strategy, and cross-border IP practice —written for attorneys, patent agents, and corporate IP counsel.
The Enlarged Board rejects a purely formal requirement to make the description mirror amended claims—but confirms that legally significant inconsistencies must still be removed.
As repeatable patent work becomes easier to standardize, the IP market is splitting into two races: efficiency in routine work, and paid trust in professional judgment.
A complete FTO is not just a patent search. It is an eight-step workflow that keeps product facts, patent rights, legal status, claim mapping, and business decisions aligned.
Comparison can quietly turn a good life into an inadequate one. The harder task is deciding what is enough without borrowing someone else’s measure of success.
The important shift is not simply better legal drafting. Astra for Law points toward AI systems that connect authoritative sources, firm knowledge, repeatable workflows, and governance.
After 24 years in patent practice, I no longer think more clients and more matters automatically mean better work. The harder constraint is how much work can still receive good judgment.
A lip implant and an artist’s blending stump looked remarkably similar. The Federal Circuit still reversed the §102 rejection because design-patent novelty is not a comparison of abstract shapes.
As patent AI gets better at producing polished answers, the scarce skill shifts toward expert evaluation: spotting the one hidden defect that can decide an OA, translation, FTO, or claim analysis.
About 90% of the accused oxidation reaction fell within the claimed pH range. Non-infringement still followed because the intrinsic record required that range to be maintained throughout the step.
G 1/25 rejects automatic description rewriting. Adaptation is required only where an inconsistency with amended claims has legal significance under the EPC.
A three-layer model can help Chinese IP firms preserve client relationships, give foreign counsel cleaner technical inputs, and build more durable China-origin patent partnerships.
Wikipedia, YouTube, forums, archived web pages, and other public sources can qualify as prior art. The real fight is usually over timing, public accessibility, what was actually disclosed, and whether it supports §102 or §103.
When expert reasoning itself can become reusable training or evaluation data, AI-era hiring raises a new question: where should candidate assessment end and professional data production begin?
A disciplined first-pass FTO screen can reduce cost and surface risk early—but only if product decomposition, claim mapping, legal-status checks, and escalation are kept separate from a formal legal opinion.
Patent practitioners can face public discipline for unverified AI citations. Using the USPTO as a case study, this article asks what accountability, correction, and quality-review mechanisms apply when patent examiners get the underlying source wrong.
The USITC’s battery investigation shows why product FTO should follow the real supply chain—from critical components and subassemblies to finished products and importers.
The Federal Circuit’s August 2026 Nielsen v. TVision decision shows why “different field” alone does not defeat §103—and why the way the inventor’s problem is defined can decide whether a reference is analogous art.
A 100-page FTO report can still leave engineers asking what to change. Useful FTO work connects each risk to product facts, claim limitations, missing information, and concrete next actions.
The near-term AI disruption in patent practice is less about eliminating experienced professionals than compressing the value of work that can be standardized, generated, and checked at scale.
A responsible FTO rarely says “no risk.” Its real value is to make patent risk visible, distinguish legal exposure from commercial exposure, and show management which risks can be designed around, licensed, monitored, or consciously accepted.
Accurate translation is not the same as U.S. filing readiness. Software and AI applications often need pre-filing adaptation for §101, §112, claim support, and future prosecution flexibility.
A missing component can defeat literal infringement of a particular claim, but a defensible design-around also requires claim construction, equivalents, prosecution history, other claims, and product evidence.
AI can compress legal work, but hourly billing still tracks actual time. The harder question is how patent professionals should price expertise, review, and accountable outcomes.
FTO should not be one-size-fits-all. A practical framework helps companies match search and analysis depth to product value, market exposure, design stage, supply chain, and decision risk.
Patentability and FTO searches may use the same databases, but they answer different legal questions. A patentable improvement can still infringe a broader earlier patent.
Under §112(a), volume is not the same as support. Written description and enablement turn on what the specification actually teaches relative to the scope of the claims.
Calling too much technology “prior art” or “conventional” can create admissions that later shape U.S. examination and validity disputes.
A stronger §103 response tests the examiner’s reasoning, compatibility, expected success, and claim gaps—not just motivation to combine.
Literal translation is not itself a §112 violation. The problem begins when grammar, antecedent basis, inconsistent naming, or unclear relationships make the claim boundaries hard to determine.
A §103 rejection needs more than matched claim elements. The examiner must connect the prior-art facts to an articulated obviousness rationale.
Examiner context can improve prosecution strategy, but it should guide communication and process—not replace claim, prior-art, and legal analysis.
First-action allowance can be excellent news, but it should still trigger a claim, file-history, continuation, IDS, and issue-readiness audit.
A faithful translation can still leave a U.S. filing strategically brittle. Adaptation reviews claim architecture, disclosure support, terminology, priority, and amendment room before the record is fixed.
AI can make invention disclosures cleaner and faster while increasing the need to verify technical provenance, inventorship, §112 support, and confidentiality.
The strongest patent-AI workflows use models for retrieval, comparison, drafting assistance, and QA—while keeping legal conclusions and professional responsibility with humans.
For Chinese applicants, IDS practice can feel like paying counsel to hand the examiner rejection material. The real discipline is earlier disclosure, smarter batching, duplicate control, and internal tracking before litigation risk appears.
The Federal Circuit’s Enviro Tech decision shows how ordinary approximation language can invalidate claims when an amended numerical boundary is not anchored in the specification or prosecution history.
Bad-faith trademark filings target Chinese Amazon sellers through brand-name hijacking, keyword squatting, and insider schemes. A practical guide to TTAB remedies, preemptive filing, and Amazon appeal strategies.
When patent prosecution fails through missed deadlines, translation errors, or inadequate searches, the liability chain between Chinese applicants, intermediary firms, and US agents becomes critical. Examining the legal framework and risk allocation strategies.
A practitioner’s roadmap for Chinese enterprises: portfolio audits, PCT vs. Paris Convention routes, multi-jurisdictional claim strategy, realistic budget planning, and practitioner selection criteria.
Effective July 20, 2026: foreign patent applicants must be represented by a registered U.S. practitioner for required submissions and post-filing activity. Full breakdown of requirements, the ADS trap, defined exceptions, cost data, and what it means for Chinese firms.
The Federal Circuit reversed on all three disputed terms in Dürr v. EFC. What the decision teaches about approximation modifiers, structural relationship terms, prosecution disclaimer scope, and claim differentiation —in a single case.
The USPTO PIER Pilot requires PCT national-phase applicants to make an early election. This guide explains the options, timing, PTA implications, and practical response issues.
Most practitioners treat the abstract as an afterthought. We argue that in cross-border prosecution —particularly when the same family is filed at both USPTO and EPO —the abstract's language can have downstream consequences for claim interpretation and prosecution history estoppel.
New matter rejections remain among the most technically demanding challenges in prosecution. This guide walks through the doctrinal framework from Ariad v. Eli Lilly and MPEP § 2163, with practical strategies for responding when the examiner has drawn a narrow interpretation of the original disclosure.
Reference numerals are common and expressly non-limiting in Chinese claims. U.S. practice also permits them—so why do U.S. practitioners often remove them before filing?
A technically accurate translation can still fail in prosecution if the translator lacks familiarity with US claims practice. We examine what distinguishes legal-grade translation from language conversion —and why Chinese patent applicants need practitioners who understand both systems.
A practical §101 response ties the examiner’s analysis, the claimed technological improvement, and the specification into one coherent eligibility story.
A narrowing amendment can overcome an examiner and later restrict the doctrine of equivalents. Festo explains what was surrendered and why the prosecution record matters.
A Final Office Action doesn't have to mean the end of the road. This article compares the strategic tradeoffs between filing an RCE, an After-Final Amendment, a Pre-Appeal Brief Request, and a Notice of Appeal —helping practitioners choose the right path based on rejection type and claim scope goals.
CNIPA’s three-step inventive-step method and U.S. Graham/KSR obviousness analysis address a related question through different structures—with real consequences for drafting and OA strategy.
“Characterized in that” is a normal feature of Chinese and PCT two-part claims. In U.S. practice, however, preserving that structure can move the claim toward Jepson territory and create record consequences.
“Configured to” can require a real functional configuration rather than mere capability—but its limiting effect depends on claim context and disclosure.
EPO oppositions are won or lost on the quality of prior art selection and the framing of inventive step arguments under the Problem-Solution Approach. This article outlines how to systematically build a prior art portfolio and structure Art. 56 EPC arguments that survive oral proceedings.
In U.S. claims, “a” usually introduces an element and “the” or “said” refers back to it. The convention is simple, but literal translation and terminology drift can make the scope unclear.
A software claim that clears CNIPA examination can still face a very different U.S. eligibility and disclosure analysis. The right response is not literal translation, but pre-filing legal-technical adaptation.
AI weakens the link between professional value and hours spent. Patent practices can respond by pricing defined outcomes, expert judgment, verification, and accountability.
The USPTO’s revised AI-assisted inventorship guidance confirms that AI is a tool, not an inventor. What patent teams should document, verify, and preserve.
Recent AI-related court errors show why legal AI needs human verification, source checking, domain rubrics, and accountable review—not blind trust in fluent output.
The USPTO’s ASAP! automated prior-art search pilot closed in June 2026. Its design still offers useful lessons about earlier prior-art visibility and prosecution strategy.
ODP is not just a terminal-disclaimer formality. Cellect, Allergan, and 2026’s Ex parte Baurin make claim, term, and family strategy central.
Global-Tech requires knowledge for induced infringement and recognizes willful blindness only when high-probability belief is paired with deliberate avoidance of the facts.
Halo ended Seagate’s rigid objective gate, while §298 bars adverse inferences from not obtaining an opinion. Opinions still matter—but as evidence and decision tools, not automatic shields.
Restriction affects election, traverse rights, withdrawn claims, rejoinder, divisionals, and later family strategy—not just which claims get examined first.
eBay rejected automatic patent injunctions and categorical denials alike. Permanent injunctions require the traditional four-factor equitable test.
Anticipation requires a legally available single reference that discloses every limitation, expressly or inherently, in the arrangement the claim requires.
Lululemon's lawsuit against Costco —design patents, registered trademarks, and trade dress combined in a single action —is a textbook case study in layered IP strategy. Direct lessons for Chinese brands building defensible market positions in the US.
Phillips remains the core U.S. claim-construction framework: read claim language through the specification and prosecution history before leaning on dictionaries or experts.
The PCT framework offers strategic leverage that many Chinese applicants underutilize. This article examines how to use the International Search Report, IPRP, and voluntary amendments under Rule 161 to shape claim scope before entering the US and European national phases.
CrossVision IP also supports Patent/IP subject-matter expert engagements, Chinese Patent Law and CNIPA analysis, AI evaluation and cross-border patent projects.