Federated Learning System Patent Landscape 2026
Federated Learning System Patent Landscape in 2026
The federated learning systems field has expanded substantially on a multi-year basis, with the top five filers holding a meaningful but not overwhelming share of the hundred largest filers’ combined output. Huawei leads by patent family count, followed closely by Ericsson and IBM, while China and the United States dominate jurisdictional filing activity.
Huawei leads a competitive but multi-player field in federated learning
Huawei Technologies holds the top position with 92 patent families, followed by Ericsson at 85 and IBM at 69, indicating a three-way lead tier with no single dominant monopolist. The top five filers together account for 30% of the hundred largest filers’ combined output, a moderate concentration level that leaves meaningful room for challengers.
The gap between the top three and the fourth-ranked Qualcomm (54 families) is notable but not prohibitive. From rank five onward, counts fall to 24 or below, suggesting a long tail of universities, regional tech firms, and specialist entrants that collectively diversify the competitive landscape.
| # | Applicant | Patent families | Share |
|---|---|---|---|
| 1 | Huawei Technologies Co., Ltd. | 92 | |
| 2 | Telefonaktiebolaget LM Ericsson (Ericsson) | 85 | |
| 3 | IBM Corporation | 69 | |
| 4 | Qualcomm Incorporated | 54 | |
| 5 | Xidian University | 24 | |
| 6 | Vellore Institute of Technology | 24 | |
| 7 | BEIJING BAIDU NETCOM SCI & TECH CO LTD | 21 | |
| 8 | WeBank Co., Ltd. (Shenzhen) | 20 | |
| 9 | Hitachi, Ltd. | 19 | |
| 10 | BEIJING UNIV OF POSTS & TELECOMM | 18 |
| # | Applicant | Patent families | Share |
|---|---|---|---|
| 11 | Robert Bosch GmbH | 17 | |
| 12 | Tencent Technology (Shenzhen) Co., Ltd. | 17 | |
| 13 | Nanjing University OF POSTS & TELECOMM | 16 | |
| 14 | Huawei Cloud Computing Technologies Co., Ltd. | 15 | |
| 15 | Intel Corporation | 13 | |
| 16 | NEC Corporation | 13 | |
| 17 | Koninklijke Philips N.V. | 13 | |
| 18 | Dell Products L.P. | 12 | |
| 19 | NTT Inc. | 12 | |
| 20 | Soongsil University Industry Cooperation Foundation | 12 |
The presence of both large telecoms (Ericsson, Qualcomm) and enterprise IT incumbents (IBM) alongside a Chinese hardware and cloud leader (Huawei) implies that federated learning patents are being pursued as both a networking-layer capability and an enterprise AI infrastructure asset — two distinct strategic motivations within the same corpus.
The most recent 18–24 months of filings are underrepresented due to standard patent publication lag; apparent volume in 2025–2026 should be read as a floor, not a ceiling. Longer-window growth, applicant concentration, and technology-route coverage are therefore more reliable signals than the latest-year bar alone.
Rapid multi-year growth anchored in AI computing, with telecoms and healthcare as active secondary branches
The annual filing trend reveals a field that emerged sharply from 2019 and expanded significantly through 2022, while the technology composition shows near-universal coverage under AI computing classes with a meaningful secondary tier in digital data processing and transmission.
Annual filing trend
Filings were negligible before 2019, then accelerated sharply to a 2022 peak of 335 patent families per year. Volume eased in 2023 and 2024, consistent with a field moving past its initial breakout phase, though the three-year recent window is still 109% above the prior three-year window. Figures for 2025 and 2026 are materially understated by publication lag and should not be read as a further decline.
↗ Hover for values · click a bar to ask EurekaTechnology composition by IPC class
G06N (AI computing models) dominates the corpus overwhelmingly, reflecting that federated learning is primarily classified as a machine-learning methodology. G06F (digital data processing) and H04L (digital information transmission) form a substantial secondary tier, underscoring the field’s dual identity as both an AI and a communications protocol technology. G06Q (business/commerce), H04W (wireless networks), and G16H (healthcare informatics) are active but comparatively sparse branches.
↗ Hover for values · click a bar to ask EurekaHighly cited patent families surfaced by the query
Citation-heavy patent families returned by the query. Use this section as citation context, not as a curated list of the most topic-specific patents.
Secure and robust federated learning system and me…
It is provided a federated learning system for aggregating gradient information representing a result of training an AI model in an edge device, the federated learning system comprising the edge device and a server apparatus, the training module in the edge device being configured to generate an edge switch share in which the encrypted aggregated gradient… (excerpt from the patent abstract)
| # | Patent | Citations |
|---|---|---|
| 1 | System and Method with Federated Learning Model fo… | 166 |
| 2 | Deep neural network optimization system for machin… | 149 |
| 3 | Anonymizing data for preserving privacy during use… | 133 |
| 4 | Private and federated learning | 126 |
| 5 | Systems and methods for distributed learning for w… | 102 |
| 6 | Privacy-preserving federated learning | 101 |
| 7 | System for secure federated learning | 100 |
| 8 | 纵向联邦学习系统优化方法、装置、设备及可读存储介质 | 98 |
Ranked by total forward citations. Citation counts favour older and broadly cited patent families, and broad or adjacent patents may appear when they match the search scope. Treat this section as citation context, not as a curated list of the most topic-specific patents. Some patent titles may be shown in their original, non-English language where an accurate translation could not be guaranteed.
What the competitive structure means for R&D prioritization
The field is in a growth stage with easing annual peaks, moderate concentration among the top filers, and active cross-sector collaboration. Engineers entering now face established incumbents but also visible adjacent branches with lower filing density.
Growth stage, past the 2022 filing peak but still expanding on a multi-year basis
The lifecycle evidence places federated learning systems firmly in the Growth stage: the recent three-year filing window is 109% above the prior three-year window, confirming sustained expansion. Annual volume has, however, eased from its 2022 peak of 335 patent families. New entrants still have room to establish positions, particularly in application-layer branches that remain sparse, but core algorithm claims are increasingly dense.
Growth · past-peak annualModerate top-tier concentration with a long challenger tail
The top five filers hold 30% of the hundred largest filers’ combined total, a level that signals competitive but not locked-up territory. The top three (Huawei, Ericsson, IBM) have meaningfully more families than rank four onward, creating a visible but crossable tier gap. The long tail — universities, regional tech companies, and fintech specialists — provides entry-point examples for niche positioning.
Moderate concentrationIBM–Rensselaer Polytechnic is the most active co-filing pair; Huawei co-files with Tsinghua University
The most active collaboration pair is IBM with Rensselaer Polytechnic Institute, with 10 co-filed patent families, suggesting IBM is using academic partnership to extend its federated learning portfolio. IBM also co-files with its own Chinese subsidiary (5 families). Huawei co-files with Tsinghua University (2 families). WeBank and Tencent have a joint filing, indicating fintech-sector collaboration. Beijing University of Posts and Telecommunications co-files with three State Grid affiliates, pointing to utility-sector application work.
Industry–academia dominantChina leads on filing volume; the US is a close second; India is a significant third
China is the leading filing jurisdiction, followed by the United States. India ranks third, ahead of EPO and WIPO PCT filings, reflecting strong university and technology-institute activity from Indian institutions. South Korea, Japan, and Germany are active at a lower level. The PCT route suggests applicants are pursuing broad international protection for core platform inventions.
China · US · India top-3Go beyond the landscape: Eureka’s TRIZ Solution agent breaks down an R&D problem and returns patented concept solutions, each with a technical approach and cited patent & literature evidence.
| Applicant | Collaborator | Co-filings |
|---|---|---|
| IBM Corporation | Rensselaer Polytechnic Institute | 10 |
| IBM Corporation | IBM China Co., Ltd. | 5 |
| Huawei Technologies Co., Ltd. | Tsinghua University | 2 |
| IBM Corporation | IBM UNITED KINGDOM LTD INTELLECTUAL PROPERTY DEPAR… | 1 |
| IBM Corporation | IBM DEUTSCHLAND GMBH | 1 |
| Xidian University | Xidian University Engineering Technology Research Institute Co., Ltd. | 1 |
| WeBank Co., Ltd. (Shenzhen) | Tencent Technology (Shenzhen) Co., Ltd. | 1 |
| Beijing University of Posts and Telecommunications | State Grid Digital Technology Holdings Co., Ltd. | 1 |
| Beijing University of Posts and Telecommunications | State Grid Jibei Electric Power Co., Ltd. Information & Communication Branch | 1 |
| Beijing University of Posts and Telecommunications | State Grid Information & Communication Yili Technology Co., Ltd. | 1 |
Co-filing pairs, ranked by the number of jointly-filed patent families.
Huawei accelerates while Ericsson and IBM pull back; Qualcomm accelerates strongly
The leader tier shows divergent momentum: Huawei and Qualcomm are growing their federated learning portfolios, while Ericsson and IBM show declining recent activity. Technology focus across all four leaders centers on G06N AI computing, with differentiation at the sub-class and application level.
Huawei Technologies
Huawei leads with 92 patent families and shows strong momentum at +81% in its recent filing window versus the prior period. Its technology emphasis is concentrated in G06N 20 (AI model computing) and G06N 3 (neural networks), with a secondary presence in G06N 5. The combination of volume leadership and accelerating recent activity positions Huawei as the most aggressive builder in this space among the top tier.
families: 92Qualcomm
Qualcomm ranks fourth overall with 54 patent families but is accelerating at +60% in its recent window, closing the gap with the top three. Its primary focus is G06N 3 (neural networks), with secondary coverage in G06N 20 and H04W 72 (wireless networks) — the wireless angle differentiating it from pure-software peers and reflecting federated learning’s role in on-device and edge inference over cellular networks.
families: 54| Applicant | Recent (3 yrs) | Trend |
|---|---|---|
| Huawei Technologies Co., Ltd. | 56 | ▲ +81% |
| Telefonaktiebolaget LM Ericsson (Ericsson) | 29 | ▼ -47% |
| IBM Corporation | 20 | ▼ -62% |
| Qualcomm Incorporated | 32 | ▲ +60% |
| Xidian University | 17 | ▲ new entrant |
| Beijing Baidu Netcom Science & Technology Co., Ltd. | 13 | ▲ new entrant |
| Hitachi, Ltd. | 19 | ▲ new entrant |
| Beijing University of Posts and Telecommunications | 11 | ▲ new entrant |
Under-served branches in healthcare informatics, wireless networks, and business applications
Beyond the dominant G06N core, several IPC branches show plausible technical relevance to federated learning but comparatively low filing density, representing areas where differentiated positioning may be achievable.
G16H · Healthcare Informatics
Healthcare informatics accounts for only 93 patent records — 3% of the top-branch share — despite federated learning’s well-established role in enabling privacy-preserving training across hospital datasets. The regulatory pressure to avoid centralizing patient data makes federated approaches technically attractive in this domain. Entry paths include partnering with health systems or device makers (Siemens Healthineers is already present at the margins of the corpus) to build application-layer claims around clinical model aggregation and differential privacy in medical AI.
Search this in Eureka →H04W · Wireless Communication Networks
H04W (wireless networks) holds 97 patent records — also 3% of the top-branch share — despite the strong alignment between federated learning and 5G/6G over-the-air model aggregation and edge inference. Qualcomm’s presence in H04W 72 shows the branch is technically validated but underpopulated relative to the AI-layer classes. Researchers targeting communication-efficient federated learning protocols, gradient compression for air-interface transmission, or network-slicing-aware aggregation strategies may find this branch more open than the G06N core.
Search this in Eureka →How leaders differ by technology route across IPC branches
Strength of each leader across the main technology routes.
| Player | G06N 20 · Computing based on AI models | G06N 3 · Computing based on AI models | G06F 21 · Electric digital data processing | H04L 9 · Digital information transmission | G06F 18 · Electric digital data processing |
|---|---|---|---|---|---|
| Telefonaktiebolaget LM Ericsson (Ericsson) | Strong · 45 | Strong · 69 | Absent | Absent | Absent |
| IBM Corporation | Strong · 59 | Moderate · 26 | Emerging · 7 | Emerging · 6 | Emerging · 4 |
| Huawei Technologies Co., Ltd. | Strong · 59 | Strong · 36 | Absent | Absent | Absent |
| Qualcomm Incorporated | Moderate · 13 | Strong · 46 | Absent | Absent | Absent |
| Xidian University | Strong · 11 | Strong · 18 | Moderate · 7 | Moderate · 7 | Moderate · 4 |
| Vellore Institute of Technology | Strong · 15 | Strong · 19 | Moderate · 6 | Absent | Absent |
| WeBank Co., Ltd. (Shenzhen) | Strong · 18 | Absent | Absent | Absent | Absent |
Frequently asked questions
The corpus in scope contains 1,708 patent families covering federated learning systems on a global basis.
Huawei Technologies is the largest filer with 92 patent families, followed by Ericsson at 85 and IBM at 69.
Yes. The field is in a Growth lifecycle stage: the recent three-year filing window is 109% above the prior three-year window. Annual volume eased from a 2022 peak, but the multi-year trajectory remains strongly upward. Recent years are further understated by publication lag.
China leads as the primary filing jurisdiction, followed by the United States and India. Europe (EPO) and WIPO PCT are the next most active, with South Korea, Japan, and Germany at lower but significant levels.
G06N (AI computing models) is by far the dominant branch. G06F (digital data processing) and H04L (digital information transmission) form a secondary tier. G06Q, H04W, and G16H are active but comparatively sparse branches.
IBM and Rensselaer Polytechnic Institute are the most active co-filing pair with 10 jointly filed patent families. IBM also co-files with its own Chinese subsidiary. Huawei co-files with Tsinghua University, and WeBank co-files with Tencent — the latter indicating fintech-sector collaboration in federated learning applications.
Ready to map your own federated learning patent landscape?
Join 18,000+ innovators using PatSnap Eureka to map any technology landscape: search 2B+ patents and papers, surface key assignees, and generate a report like this in minutes.
Disclaimer. This page is generated from PatSnap Eureka data drawn from a limited snapshot of global patent and scientific-literature records, and is provided for general information and reference only.
Patent data carries inherent limitations: recent filings (typically the most recent 18–24 months) are under-counted due to standard publication lag; counts may be reported at either a patent-family or a patent-record basis and are not always directly comparable; classification, applicant-name, and citation data may contain errors, duplicates, or omissions; and the underlying search query defines and constrains the scope shown. As a result, the analysis may be incomplete or inaccurate and may not reflect the full technology landscape.
Nothing on this page constitutes an exhaustive prior-art, novelty, freedom-to-operate, or validity search, nor does it constitute legal, financial, investment, or professional advice, and it should not be relied upon as such. Any patent, commercial, or strategic decision should be verified independently and reviewed with qualified patent, legal, and domain professionals. PatSnap makes no warranties, express or implied, as to the accuracy, completeness, or fitness for any particular purpose of the information presented.