About
Patent AI

Patent AI

Tracking Patent Ai legal and regulatory developments.

2 entries in Litigator Tracker

LawSnap Briefing Updated October 7, 2026

State of play.

  • The patent AI tooling market has crossed from experimentation to competition. DeepIP launched its prosecution copilot out of beta with Microsoft Word integration, LexisNexis introduced Protégé in PatentSight+, and industry data cited by Law360 shows AI adoption in IP practice reached 85 percent in 2025—up from 57 percent in 2023—with startups including Patlytics and DeepIP raising $14M and $25M respectively in recent rounds (→ AI Legal-Tech Startups Target Patent Practice as Big Vendors Circle).
  • The USPTO has hardened foreign applicant requirements while domestic prosecution reform continues. The ban on pro se patent prosecution by foreign-domiciled applicants took effect July 20, 2026, with no grandfathering—retroactively capturing pending applications—while the former USPTO deputy commissioner for AI has moved to Jones Day, signaling where market demand for government-side expertise is flowing (→ USPTO ends self-representation for foreign-domiciled patent applicants).
  • China is executing a coordinated IP strategy across filing volume, examination access, and portfolio quality metrics. WIPO data shows China accounted for roughly 70 percent of global generative AI patent family publications from 2014 to 2023, CNIPA has finalized priority examination rules that exclude foreign applicants from direct access, and CNIPA has redefined "high-value patent" to emphasize strategic emerging industries and international scope over financing metrics .
  • The UK Supreme Court has signaled that patent pools cannot shield SEP holders from FRAND obligations. Tesla's revival of its lawsuit against InterDigital and Avanci establishes that pool membership does not extinguish FRAND licensing duties—a principle with implications for connected vehicle and IoT licensing structures well beyond the automotive sector .
  • For counsel advising technology companies, healthcare AI developers, or cross-border IP portfolios, the practical baseline is a convergence of four pressures: AI tooling is reshaping prosecution economics, USPTO procedural reform and the foreign pro se ban are generating immediate compliance obligations, China's asymmetric enforcement and examination access are widening competitive exposure, and FRAND doctrine is being tested in a new judicial forum.

Where things stand.

  • USPTO has affirmatively embraced AI patentability while requiring significant human contribution. Director Squires' guidance confirms AI inventions are patentable; the "significant human contribution" standard governs inventorship, with AI tools treated as analogous to laboratory equipment rather than disqualifying co-inventors .
  • Design patent coverage has expanded to virtual, AR, and projected interfaces. The USPTO's March 2026 supplemental guidance under 35 U.S.C. § 171 eliminates the requirement to depict a physical display panel and applies retroactively to all pending applications .
  • The foreign pro se ban is now in force with no transition relief. All patent applicants and owners domiciled outside the US or its territories must retain a registered US patent practitioner—covering both new filings and pending applications, and capturing any multi-applicant case with even one foreign-domiciled party (→ USPTO ends self-representation for foreign-domiciled patent applicants).
  • PTAB institution decisions are final and unreviewable on appeal. The FedEx v. Qualcomm precedential decision holds that RPI challenges under § 312(a)(2) are integral to the institution decision and blocked from appellate review under § 314(d), with narrow exceptions only for constitutional violations or actions plainly outside statutory authority .
  • Equitable estoppel now requires proof of actual reliance, not mere silence. The Federal Circuit's reversal in Fraunhofer v. Sirius XM holds that a patentee's five-year silence constitutes misleading conduct but does not establish estoppel unless the defendant proves its business decisions were actually influenced by that silence .
  • Ex parte reexamination is displacing IPR as the preferred patent challenge vehicle, with new pre-order procedures and Director gatekeeping reshaping the challenge landscape; parallel reexamination proceedings have become a material preliminary injunction variable in district court .
  • Inventorship is the sharpest litigation risk in the AI-assisted invention era. The Federal Circuit's precedential ruling in Fortress Iron v. Digger Specialties establishes that an uncorrectable inventorship omission invalidates a patent outright, with no AIA savings clause available .
  • Healthcare AI developers face a structural three-front compliance challenge. Patent counsel, FDA regulators, and HIPAA compliance teams must align before first commercial release—FDA locks in product specifications, patent law requires documented human inventorship, and HIPAA restricts training data use in ways that constrain model monitoring and retraining .
  • China's IP enforcement posture has hardened materially. The SPC's IP Court saw a 66.7% year-over-year increase in punitive damages cases in 2025, foreign-related filings surged 34.1%, and CNIPA's revised priority examination rules exclude foreign applicants from direct expedited review—raising unresolved TRIPS national-treatment questions .
  • The UK Supreme Court has revived Tesla's FRAND challenge to the Avanci 5G pool, holding that pool membership does not extinguish FRAND licensing obligations—a ruling that will shape how UK courts treat SEP disputes in the automotive and IoT sectors .
  • Patent disclosure irrevocably forfeits trade secret status. International Medical Devices v. Cornell confirms that information disclosed in publicly available patents is "generally known" under California's UTSA, eliminating any concurrent trade secret claim .
  • The CRISPR priority dispute remains unresolved, with the PTAB's remand reaffirmation of Broad Institute priority likely to return to the Federal Circuit—billions in gene-editing licensing value remain at stake .

Latest developments.

Active questions and open splits.

  • What constitutes "significant human contribution" in AI-assisted invention? The USPTO treats AI tools as laboratory equipment, but no Federal Circuit decision has tested the standard in litigation—leaving inventorship documentation requirements, and the line between AI-assisted and AI-generated work, unsettled for prosecution and due diligence .
  • How should IPR petitioners recalibrate RPI strategy after FedEx v. Qualcomm? With appellate review of RPI determinations foreclosed, petitioners must resolve RPI exposure before filing—and patent owners must press RPI objections forcefully pre-institution knowing the PTAB's handling will not be corrected on appeal .
  • Does CNIPA's foreign-applicant exclusion from priority examination violate TRIPS national-treatment obligations? The criticism accompanied the draft and survived the comment period unchanged—but no formal dispute mechanism has been invoked, leaving the question open for companies that depend on accelerated grant timelines in China .
  • Will the UK Supreme Court's Tesla ruling reshape SEP pool licensing structures globally? The holding that pool membership does not extinguish FRAND obligations applies directly to the Avanci model—but how the High Court defines the scope of Tesla's remand proceedings, and whether other jurisdictions follow, is unresolved .
  • What reliance evidence survives the tightened equitable estoppel standard after Fraunhofer v. Sirius XM? The actual-influence requirement means defendants in collaborative or licensing relationships must document that a patentee's silence—not independent market forces—drove their technology decisions; the remand will test what evidence satisfies that burden at trial .
  • Which AI patent tooling incumbents will acquire or partner with which startups—and will that consolidation determine prosecution workflow standards? The market is moving from experimentation to competition, but the partnership discussions disclosed by startup CEOs remain undisclosed in their terms; the acquisition or partnership landscape will likely determine which tools become standard across Am Law firms (→ AI Legal-Tech Startups Target Patent Practice as Big Vendors Circle).
  • What is the § 101 floor for software and AI-adjacent patents after the USPTO's pro-AI guidance? District courts continue granting early dismissals of patents covering digitized business processes while the USPTO affirms AI inventions are patentable—the gap between prosecution-stage eligibility and litigation-stage § 101 challenges remains wide .

What to watch.

  • USPTO enforcement mechanics for the foreign pro se ban—how the agency handles pending applications with foreign-domiciled parties who have not yet retained counsel, and whether any transition guidance issues.
  • High Court proceedings on remand in Tesla v. InterDigital/Avanci—the scope of what the court will consider on FRAND terms will define how pool-based SEP licensing structures operate in the UK and signal to other jurisdictions.
  • SXM's reliance case on remand in Fraunhofer v. Sirius XM—the evidence SXM marshals to prove actual reliance will define the practical boundaries of equitable estoppel in technology partnership and licensing scenarios.
  • Whether major legal-tech incumbents announce acquisitions or formal partnerships with DeepIP, Patlytics, Solve Intelligence, or PatSnap—those deals will determine which AI prosecution tools become workflow standards across large IP practices.
  • Early cases applying China's revised punitive damages framework and CNIPA's new high-value patent definition—the first enforcement actions will signal how aggressively courts deploy escalated awards against foreign parties and how the redefined metric reshapes prosecution priorities in strategic emerging industries.
  • Federal Circuit appeal in the CRISPR interference—whether the court accepts the PTAB's swift remand reaffirmation or imposes a more demanding conception standard will reshape biotech licensing strategy across gene-editing portfolios.

2 Contributing Entries

UN releases 2026 International AI Safety Report warning of enormous benefits and existential risks

The United Nations released the International AI Safety Report 2026, a comprehensive assessment concluding that advanced artificial intelligence presents both transformative opportunities and escalating dangers. The report, led by the UN agency for digital technology, finds that AI can accelerate development in health, education, and financial services in developing nations while simultaneously enabling cyberattacks, deepfake fraud, non-consensual intimate imagery, and biological weapon design. The core finding: AI capabilities in critical fields like biological research are advancing faster than governance frameworks, creating a dangerous gap between what is technologically possible and what remains safe.

mail Subscribe to Patent AI email updates

Primary sources. No fluff. Straight to your inbox.

Also on LawSnap