China Trademark Law - Search Obligations, Evidence of Use, and Maintenance of Registrations - An Analysis of the “Huawei Cloud Yaoguang” Cancellation Review Case
Published 7 June 2026
Sarah Xuan
On May 26, 2025, CNIPA issued the Guidelines for Filing Applications to Cancel Registered Trademarks Not Used for Three Consecutive Years, imposing clearer requirements regarding preliminary investigations and the authenticity of supporting materials at the application stage of non-use cancellation proceedings. The Guidelines did not alter the statutory rule under Article 49(2) of the Trademark Law concerning cancellation for non-use for three consecutive years, but they increased the requirements imposed on applicants with respect to investigation, explanation, and preparation of supporting materials when filing such applications.
The cancellation review case concerning the “Huawei Cloud Yaoguang” trademark, Registration No. 43736863, arose shortly after this adjustment in filing practice. In that case, the original cancellation applicant, Tianjin Baisihui Technology Co., Ltd., asserted that, “after extensive searches, it found no actual use of the mark under review.” In its original non-use cancellation decision, CNIPA found that part of the evidence of use submitted by Huawei Technologies Co., Ltd. was valid, maintained the registration of the mark under review for goods including “recorded computer operating programs; computer software, recorded; downloadable computer application software; downloadable mobile-phone application software; wearable computers; smartwatches (data processing); tablet computers; recorded or downloadable computer software platforms; notebook computers; computers; central processing units (CPUs); integrated circuit cards; computer hardware; humanoid robots with artificial intelligence,” and cancelled the registration for the remaining goods, including “smartphones.” Dissatisfied with the determination maintaining the registration for certain goods, the applicant sought review. Following review, CNIPA maintained the registration of the mark under review for the goods subject to review.
The analytical significance of the case is not confined to its individual outcome. It illustrates several important relationships in non-use cancellation cases following the issuance of the Guidelines: the relationship between the search obligation at the application stage and the substantive determination of use; the relationship between ordinary online searches and industry-specific evidence of use; the relationship between goods actually used and the scope of registration maintained for similar goods; and the relationship between evidentiary challenges and substantive rebuttals in review proceedings. These relationships provide a clearer view of how the evidentiary structures of applicants and registrants unfold in non-use cancellation proceedings after implementation of the Guidelines.
I. Searches at the Application Stage Following the Non-Use Cancellation Guidelines The Guidelines require an applicant to provide evidence of a preliminary investigation at the application stage, with the aim of enhancing the authenticity and procedural regularity of non-use cancellation applications and reducing the adverse impact of arbitrary applications on the orderliness of the register and the stability of trademark rights. The applicant must demonstrate that its allegation of “non-use” is not made solely to clear a blocking mark or to test the procedure, but is based on an external search and factual verification of a certain scope.
This requirement is consistent with the fundamental function of the cancellation regime for non-use for three consecutive years. In its decision in this case, CNIPA expressly stated: “The legislative purpose of the provisions on non-use cancellation is to encourage trademark registrants to use their trademarks and enable trademarks to perform their functions, to avoid the idling and waste of trademark resources, and to remove trademarks for which there is no genuine intention to use.” This statement indicates that the central concern of the non-use cancellation regime is not merely to provide applicants with a tool for clearing blocking marks, but to require use so that registered trademarks perform their source-identifying function and to remove registrations lacking a genuine intention to use.
Accordingly, acceptance of a non-use cancellation application does not mean that the fact of “non-use” has been established. The function of the investigative materials submitted at the application stage is to demonstrate that the cancellation application has a prima facie factual basis; substantive examination of a non-use cancellation case still depends on whether the registrant can submit evidence of genuine, public, lawful, and valid commercial use during the prescribed period. The two operate at different procedural stages and perform different institutional functions.
The “Huawei Cloud Yaoguang” case illustrates this point. The applicant asserted that extensive searches had revealed no actual use of the mark under review and sought cancellation on that basis. The fact that the application proceeded to the non-use cancellation procedure indicates that the application materials met the requirements for initiating the procedure in form and at the prima facie factual level. During the subsequent examination, however, the respondent submitted materials including content from its official website, media reports, search results, product descriptions, and brand honors, which were sufficient for CNIPA to find valid commercial use of the mark under review during the prescribed period. A finding of “no use discovered” at the application stage does not preclude the registrant from possessing and submitting other evidence of use.
II. Industry Characteristics and Probative Value of Evidence of Use for Technology Trademarks In this case, the principal evidence submitted by the respondent included promotional reports concerning “Huawei Cloud Yaoguang” and honors received by its algorithm team from sources including Baidu, the 51CTO Blog, Shangyun Wuyou, Sohu, China Youth Online, China Daily, and the respondent’s official website, as well as product search results and descriptions on platforms including Zhihu and Baidu. CNIPA also obtained materials concerning use of the “Huawei Cloud Yaoguang” brand and honors received, which the respondent had submitted in the original non-use cancellation proceedings.
The foregoing evidence differs markedly from the packaging photographs, sales invoices, distribution agreements, or e-commerce orders commonly seen in cases involving traditional consumer goods. Its form is more consistent with the commercial communication practices applicable to cloud computing, software platforms, and enterprise technology products. Trademark use for such goods is often reflected in official website product descriptions, technical solutions, industry reports, developer communities, media communications, awards and honors, and ecosystem partnerships. The transaction counterparties, business models, and communication channels for such goods are highly industry-specific, and evidence of use need to therefore be assessed in light of the corresponding commercial practices.
In this regard, CNIPA expressly stated in its decision: “When examining evidence of trademark use, full consideration shall be given to general commercial practices in market operations and the industry characteristics of business conduct, with particular attention paid to the trademark registrant’s genuine intention to use and the manner in which the mark under review performs a source-identifying function in its business activities.” This statement reflects that, when examining evidence of use for technology trademarks, the focus needs to be on determining whether the relevant sign has been associated with specific goods in genuine commercial activities and performs the function of identifying the source of those goods among the relevant public.
In this case, CNIPA found that the evidence submitted by the respondent, including “promotional reports concerning ‘Huawei Cloud Yaoguang’ from media such as Zhihu, Baidu, the 51CTO Blog, Shangyun Wuyou, Sohu, China Youth Online, China Daily, and the respondent’s official website,” was capable of proving that “Huawei Cloud Yaoguang” had been used for cloud operating systems during the prescribed period and “had become known to the relevant public and was capable of distinguishing the source of the goods.” The decision further stated that the relevant evidence “reflected such elements as the mark under review, the cloud operating systems and other goods actually used, the relevant time, and the identity of the respondent,” and on that basis found that the mark under review had been put to “public, genuine, and lawful commercial use” for cloud operating systems and other goods.
It follows that the probative value of media reports, search results, and official website descriptions need to be assessed in the context of the complete evidentiary chain. A single media report will ordinarily be insufficient, standing alone, to prove trademark use. Where the relevant materials corroborate one another and correspond to the prescribed period, the user, the mark under review, the goods actually used, and the commercial activities, they may collectively prove genuine commercial use of the trademark and its source-identifying function.
From the perspective of evidentiary structure, evidence of use in non-use cancellation cases generally must demonstrate the following elements: first, the evidence must have been created within the prescribed period; second, the user must be attributable to the trademark registrant, a licensee, or another party whose use is not contrary to the registrant’s will; third, the sign actually used must be the registered trademark or be substantially identical to it in terms of identification; fourth, the use must relate to the designated goods or goods similar thereto; fifth, the relevant conduct must occur in advertising, transactions involving goods, exhibitions or displays, platform promotion, or other commercial activities; and sixth, the trademark must perform a source-identifying function in the relevant business activities.
This method of assessing evidence has direct implications both for the applicant’s search obligation at the application stage and for its subsequent evidentiary challenges. In relation to technology trademarks, the applicant’s search need to not be confined to traditional sales channels, but need to also cover the registrant’s official website, industry media, technical communities, product descriptions, developer platforms, and other public channels suited to the industry’s commercial characteristics. At the evidentiary challenge stage, the applicant likewise needs to not reject the probative value of materials merely because they consist of media reports or search results, but need to raise specific objections concerning the time, user, sign, goods, commercial use, and source-identifying function reflected in the evidence. Untargeted procedural objections will ordinarily be insufficient to undermine a body of evidence that has already formed a complete and mutually corroborative chain.
III. Scope of Registration Maintained for Similar Goods Another noteworthy issue in this case is the similarity between the goods actually used and the designated goods. CNIPA reiterated in its decision: “Where a trademark registrant or licensee uses a registered trademark on one of the designated goods, the registration may be maintained for goods similar to that good.” This rule determines the scope of registration maintained in non-use cancellation cases and requires the examination to go beyond the question of “whether use exists” and further determine “which designated goods may be covered by the goods actually used.”
In this case, CNIPA found that the evidence submitted by the respondent was capable of proving genuine and valid commercial use of the mark under review for cloud operating systems and other goods. It further found that all of the goods subject to review and designated for use under the mark, including “recorded computer operating programs,” were the same as or similar to the goods actually used. Accordingly, the registration of the mark under review was to be maintained for the goods subject to review.
This finding concerns a rule that is common in non-use cancellation cases but easily underestimated: where a registrant uses a registered trademark on one type of goods, the registration may be maintained for goods similar thereto. The non-use cancellation regime is intended to remove registered trademarks that have remained idle for a prolonged period and for which there is no genuine intention to use; it does not require the registrant to submit wholly independent evidence of use for every designated item of goods. So long as the registrant has made genuine and valid use of the mark on certain goods during the prescribed period, and those goods are the same as or similar to several designated goods, the registration for the corresponding designated goods may be maintained.
In the “Huawei Cloud Yaoguang” case, the actual use related to goods such as “cloud operating systems,” while the maintained scope extended to numerous Class 9 goods, including computer operating programs, recorded software, downloadable application software, software platforms, computer hardware, central processing units, integrated circuit cards, and humanoid robots with artificial intelligence. This outcome demonstrates that, in the contexts of software, cloud computing, and intelligent technologies, a finding of similarity between the goods actually used and the designated goods may have a substantial spillover effect. An applicant that focuses only on whether the mark under review appears item by item on each designated good may underestimate the effect of the similar-goods rule on the scope of registration maintained.
Of course, maintenance of registration for similar goods is not without limits. The original non-use cancellation decision cancelled the registration of the mark under review for the remaining goods, including “smartphones,” demonstrating that evidence of actual use still has boundaries. Which goods may be maintained on the strength of evidence of use depends on the degree of similarity between the goods actually used and the designated goods in terms of function, purpose, consumers, sales channels, production sectors, technological connection, and market perception. For an applicant, a non-use cancellation strategy needs to not merely determine “whether use exists,” but need to further assess “if use exists, which similar goods that use may cover.” This assessment directly determines the expected benefit of the cancellation application.
IV. Limitation of the Scope of Review: Procedural Choices in Cases of Partial Cancellation and Partial Maintenance The procedural structure of this case is also worthy of analysis. In the original non-use cancellation decision, the registration of the mark under review was maintained for certain goods and cancelled for the remaining goods, including “smartphones.” CNIPA expressly stated in its decision: “The respondent in this case did not file an application for review within the prescribed period; accordingly, smartphones and the other remaining goods do not fall within the scope of review in this case.” The goods subject to review in this case were therefore limited to those for which the original decision maintained registration and in respect of which the applicant sought review.
This treatment reflects the rule limiting the scope of review in non-use cancellation proceedings. A non-use cancellation review does not automatically entail a comprehensive re-examination of the original cancellation decision. The scope of review depends on the relief sought by the party applying for review and the portion of the decision with which that party is dissatisfied. Where the applicant challenges the portion maintaining registration, the review concerns the goods for which registration was maintained; where the registrant challenges the portion cancelling registration, it must separately file an application for review in accordance with law. A portion not placed under review will ordinarily not automatically be re-examined in review proceedings initiated by the other party.
This procedural rule has practical significance for both registrants and applicants. For a registrant, where the original non-use cancellation decision has cancelled registration for certain goods and the registrant considers that cancellation improper, it need to timely seek review and may not expect the matter to be resolved in review proceedings brought by the applicant against the portion maintaining registration. For an applicant, the grounds for review and evidentiary challenges need to focus on the goods for which the original decision maintained registration, with specific arguments directed to the correspondence and similarity between those goods and the goods actually used.
V. Strategy for Challenging Evidence: The Relationship Between Procedural Objections and Substantive Rebuttals One of the applicant’s principal grounds for review was that, “after extensive searches, it found no actual use of the mark under review,” and that it had not been afforded an opportunity to examine and challenge the evidence of use submitted by the respondent; it therefore sought review and requested such an opportunity. This ground has procedural significance, but the key issue in a non-use cancellation review remains the substantive determination. An opportunity to examine and challenge evidence may assist the applicant in scrutinizing and rebutting the evidence, but it cannot substitute for a specific evidentiary analysis.
In a non-use cancellation review, an applicant seeking effectively to rebut a registrant’s evidence of use need to raise specific arguments directed to the critical links in the evidentiary chain. First, it needs to examine whether the evidence falls within the prescribed period. The prescribed period in this case was from October 29, 2021 to October 28, 2024; materials outside that period can serve only a corroborative function. Second, it needs to examine whether the evidence displays the mark under review itself, or whether the sign used is substantially identical to the mark under review in terms of identification. Third, it needs to examine whether the user is attributable to the registrant, a licensee, or another party whose use is not contrary to the registrant’s will. Fourth, it need to examine whether the goods actually reflected in the evidence are the same as or similar to the goods subject to review. Finally, it needs to examine whether the relevant use constitutes public use in commercial activities and is capable of performing a source-identifying function.
In this case, CNIPA ultimately accepted the respondent’s evidence, indicating that the applicant’s evidentiary challenges failed to undermine the overall probative value of the evidence of use with respect to time, user, sign, goods, and source-identifying function. The case demonstrates that an evidentiary challenge in a non-use cancellation review need to not stop at the procedural assertion that the evidence “was not subject to examination and challenge,” but need to address specific weaknesses in the evidentiary chain. Where the registrant’s evidentiary system already forms a relatively complete closed loop, the applicant must present more refined substantive rebuttals, particularly as to whether the goods actually used and the goods subject to review constitute similar goods.
Conclusion: Implications for Strategies to Clear Blocking Marks Through Non-Use Cancellation The “Huawei Cloud Yaoguang” case provides a concrete example of a strategy to clear blocking marks following issuance of the Non-Use Cancellation Guidelines. Non-use cancellation remains an important avenue for addressing genuinely idle trademarks. However, in relation to large technology companies, cloud-service brands, software platforms, artificial-intelligence products, or enterprise technology signs, applicants must assess the likelihood of use with greater caution. Ordinary search engines, e-commerce platforms, and corporate-status searches may form the basis of an initial investigation, but they may not be sufficient to cover the actual-use scenarios of technology trademarks.
Applicants for non-use cancellation need to expand the scope of their searches in accordance with industry characteristics. For trademarks relating to software, cloud computing, and artificial intelligence, searches need to not be confined to product sales pages, but need to also cover the registrant’s official website, product center, developer documentation, technical white papers, cloud marketplaces, industry media, technical blogs, developer communities, official press releases, awards and honors, partner pages, conferences and events, tendering and bidding information, and solution materials. For trademarks with pronounced technological or B2B characteristics, evidence of use may be found in commercial promotion, technical communications, and industry applications, and may not take the form of transaction evidence typical of ordinary consumer goods.
The case also demonstrates that an assessment of the expected benefit of a non-use cancellation application need to take account of the rule maintaining registration for similar goods. Even where an applicant determines that a trademark is used only for one specific technology product, it need to continue to analyze whether that use may cover other similar items among the designated goods. Where the goods actually used are similar to a large number of designated goods, even a successful non-use cancellation application may result only in cancellation for some goods. An applicant with a clear objective of removing a blocking mark need to, before filing, assess whether the target goods may be protected by the spillover effect of the registrant’s evidence of actual use for similar goods.
More broadly, a non-use cancellation strategy needs to be coordinated with invalidation, opposition, litigation, and administrative referral mechanisms. Where the principal defect of a blocking trademark is that it is idle and unused, non-use cancellation remains directly applicable. Where the blocking trademark also involves free-riding on another brand, conflict with prior rights, or bad-faith registration, reliance on non-use cancellation alone may not address the central defect in the case. Strategies for clearing blocking marks after 2025 place greater emphasis on an applicant’s integrated assessment of trademark status, likelihood of use, similarity of goods, and procedural objectives.
The cancellation review case concerning the “Huawei Cloud Yaoguang” trademark, Registration No. 43736863, arose shortly after this adjustment in filing practice. In that case, the original cancellation applicant, Tianjin Baisihui Technology Co., Ltd., asserted that, “after extensive searches, it found no actual use of the mark under review.” In its original non-use cancellation decision, CNIPA found that part of the evidence of use submitted by Huawei Technologies Co., Ltd. was valid, maintained the registration of the mark under review for goods including “recorded computer operating programs; computer software, recorded; downloadable computer application software; downloadable mobile-phone application software; wearable computers; smartwatches (data processing); tablet computers; recorded or downloadable computer software platforms; notebook computers; computers; central processing units (CPUs); integrated circuit cards; computer hardware; humanoid robots with artificial intelligence,” and cancelled the registration for the remaining goods, including “smartphones.” Dissatisfied with the determination maintaining the registration for certain goods, the applicant sought review. Following review, CNIPA maintained the registration of the mark under review for the goods subject to review.
The analytical significance of the case is not confined to its individual outcome. It illustrates several important relationships in non-use cancellation cases following the issuance of the Guidelines: the relationship between the search obligation at the application stage and the substantive determination of use; the relationship between ordinary online searches and industry-specific evidence of use; the relationship between goods actually used and the scope of registration maintained for similar goods; and the relationship between evidentiary challenges and substantive rebuttals in review proceedings. These relationships provide a clearer view of how the evidentiary structures of applicants and registrants unfold in non-use cancellation proceedings after implementation of the Guidelines.
I. Searches at the Application Stage Following the Non-Use Cancellation Guidelines The Guidelines require an applicant to provide evidence of a preliminary investigation at the application stage, with the aim of enhancing the authenticity and procedural regularity of non-use cancellation applications and reducing the adverse impact of arbitrary applications on the orderliness of the register and the stability of trademark rights. The applicant must demonstrate that its allegation of “non-use” is not made solely to clear a blocking mark or to test the procedure, but is based on an external search and factual verification of a certain scope.
This requirement is consistent with the fundamental function of the cancellation regime for non-use for three consecutive years. In its decision in this case, CNIPA expressly stated: “The legislative purpose of the provisions on non-use cancellation is to encourage trademark registrants to use their trademarks and enable trademarks to perform their functions, to avoid the idling and waste of trademark resources, and to remove trademarks for which there is no genuine intention to use.” This statement indicates that the central concern of the non-use cancellation regime is not merely to provide applicants with a tool for clearing blocking marks, but to require use so that registered trademarks perform their source-identifying function and to remove registrations lacking a genuine intention to use.
Accordingly, acceptance of a non-use cancellation application does not mean that the fact of “non-use” has been established. The function of the investigative materials submitted at the application stage is to demonstrate that the cancellation application has a prima facie factual basis; substantive examination of a non-use cancellation case still depends on whether the registrant can submit evidence of genuine, public, lawful, and valid commercial use during the prescribed period. The two operate at different procedural stages and perform different institutional functions.
The “Huawei Cloud Yaoguang” case illustrates this point. The applicant asserted that extensive searches had revealed no actual use of the mark under review and sought cancellation on that basis. The fact that the application proceeded to the non-use cancellation procedure indicates that the application materials met the requirements for initiating the procedure in form and at the prima facie factual level. During the subsequent examination, however, the respondent submitted materials including content from its official website, media reports, search results, product descriptions, and brand honors, which were sufficient for CNIPA to find valid commercial use of the mark under review during the prescribed period. A finding of “no use discovered” at the application stage does not preclude the registrant from possessing and submitting other evidence of use.
II. Industry Characteristics and Probative Value of Evidence of Use for Technology Trademarks In this case, the principal evidence submitted by the respondent included promotional reports concerning “Huawei Cloud Yaoguang” and honors received by its algorithm team from sources including Baidu, the 51CTO Blog, Shangyun Wuyou, Sohu, China Youth Online, China Daily, and the respondent’s official website, as well as product search results and descriptions on platforms including Zhihu and Baidu. CNIPA also obtained materials concerning use of the “Huawei Cloud Yaoguang” brand and honors received, which the respondent had submitted in the original non-use cancellation proceedings.
The foregoing evidence differs markedly from the packaging photographs, sales invoices, distribution agreements, or e-commerce orders commonly seen in cases involving traditional consumer goods. Its form is more consistent with the commercial communication practices applicable to cloud computing, software platforms, and enterprise technology products. Trademark use for such goods is often reflected in official website product descriptions, technical solutions, industry reports, developer communities, media communications, awards and honors, and ecosystem partnerships. The transaction counterparties, business models, and communication channels for such goods are highly industry-specific, and evidence of use need to therefore be assessed in light of the corresponding commercial practices.
In this regard, CNIPA expressly stated in its decision: “When examining evidence of trademark use, full consideration shall be given to general commercial practices in market operations and the industry characteristics of business conduct, with particular attention paid to the trademark registrant’s genuine intention to use and the manner in which the mark under review performs a source-identifying function in its business activities.” This statement reflects that, when examining evidence of use for technology trademarks, the focus needs to be on determining whether the relevant sign has been associated with specific goods in genuine commercial activities and performs the function of identifying the source of those goods among the relevant public.
In this case, CNIPA found that the evidence submitted by the respondent, including “promotional reports concerning ‘Huawei Cloud Yaoguang’ from media such as Zhihu, Baidu, the 51CTO Blog, Shangyun Wuyou, Sohu, China Youth Online, China Daily, and the respondent’s official website,” was capable of proving that “Huawei Cloud Yaoguang” had been used for cloud operating systems during the prescribed period and “had become known to the relevant public and was capable of distinguishing the source of the goods.” The decision further stated that the relevant evidence “reflected such elements as the mark under review, the cloud operating systems and other goods actually used, the relevant time, and the identity of the respondent,” and on that basis found that the mark under review had been put to “public, genuine, and lawful commercial use” for cloud operating systems and other goods.
It follows that the probative value of media reports, search results, and official website descriptions need to be assessed in the context of the complete evidentiary chain. A single media report will ordinarily be insufficient, standing alone, to prove trademark use. Where the relevant materials corroborate one another and correspond to the prescribed period, the user, the mark under review, the goods actually used, and the commercial activities, they may collectively prove genuine commercial use of the trademark and its source-identifying function.
From the perspective of evidentiary structure, evidence of use in non-use cancellation cases generally must demonstrate the following elements: first, the evidence must have been created within the prescribed period; second, the user must be attributable to the trademark registrant, a licensee, or another party whose use is not contrary to the registrant’s will; third, the sign actually used must be the registered trademark or be substantially identical to it in terms of identification; fourth, the use must relate to the designated goods or goods similar thereto; fifth, the relevant conduct must occur in advertising, transactions involving goods, exhibitions or displays, platform promotion, or other commercial activities; and sixth, the trademark must perform a source-identifying function in the relevant business activities.
This method of assessing evidence has direct implications both for the applicant’s search obligation at the application stage and for its subsequent evidentiary challenges. In relation to technology trademarks, the applicant’s search need to not be confined to traditional sales channels, but need to also cover the registrant’s official website, industry media, technical communities, product descriptions, developer platforms, and other public channels suited to the industry’s commercial characteristics. At the evidentiary challenge stage, the applicant likewise needs to not reject the probative value of materials merely because they consist of media reports or search results, but need to raise specific objections concerning the time, user, sign, goods, commercial use, and source-identifying function reflected in the evidence. Untargeted procedural objections will ordinarily be insufficient to undermine a body of evidence that has already formed a complete and mutually corroborative chain.
III. Scope of Registration Maintained for Similar Goods Another noteworthy issue in this case is the similarity between the goods actually used and the designated goods. CNIPA reiterated in its decision: “Where a trademark registrant or licensee uses a registered trademark on one of the designated goods, the registration may be maintained for goods similar to that good.” This rule determines the scope of registration maintained in non-use cancellation cases and requires the examination to go beyond the question of “whether use exists” and further determine “which designated goods may be covered by the goods actually used.”
In this case, CNIPA found that the evidence submitted by the respondent was capable of proving genuine and valid commercial use of the mark under review for cloud operating systems and other goods. It further found that all of the goods subject to review and designated for use under the mark, including “recorded computer operating programs,” were the same as or similar to the goods actually used. Accordingly, the registration of the mark under review was to be maintained for the goods subject to review.
This finding concerns a rule that is common in non-use cancellation cases but easily underestimated: where a registrant uses a registered trademark on one type of goods, the registration may be maintained for goods similar thereto. The non-use cancellation regime is intended to remove registered trademarks that have remained idle for a prolonged period and for which there is no genuine intention to use; it does not require the registrant to submit wholly independent evidence of use for every designated item of goods. So long as the registrant has made genuine and valid use of the mark on certain goods during the prescribed period, and those goods are the same as or similar to several designated goods, the registration for the corresponding designated goods may be maintained.
In the “Huawei Cloud Yaoguang” case, the actual use related to goods such as “cloud operating systems,” while the maintained scope extended to numerous Class 9 goods, including computer operating programs, recorded software, downloadable application software, software platforms, computer hardware, central processing units, integrated circuit cards, and humanoid robots with artificial intelligence. This outcome demonstrates that, in the contexts of software, cloud computing, and intelligent technologies, a finding of similarity between the goods actually used and the designated goods may have a substantial spillover effect. An applicant that focuses only on whether the mark under review appears item by item on each designated good may underestimate the effect of the similar-goods rule on the scope of registration maintained.
Of course, maintenance of registration for similar goods is not without limits. The original non-use cancellation decision cancelled the registration of the mark under review for the remaining goods, including “smartphones,” demonstrating that evidence of actual use still has boundaries. Which goods may be maintained on the strength of evidence of use depends on the degree of similarity between the goods actually used and the designated goods in terms of function, purpose, consumers, sales channels, production sectors, technological connection, and market perception. For an applicant, a non-use cancellation strategy needs to not merely determine “whether use exists,” but need to further assess “if use exists, which similar goods that use may cover.” This assessment directly determines the expected benefit of the cancellation application.
IV. Limitation of the Scope of Review: Procedural Choices in Cases of Partial Cancellation and Partial Maintenance The procedural structure of this case is also worthy of analysis. In the original non-use cancellation decision, the registration of the mark under review was maintained for certain goods and cancelled for the remaining goods, including “smartphones.” CNIPA expressly stated in its decision: “The respondent in this case did not file an application for review within the prescribed period; accordingly, smartphones and the other remaining goods do not fall within the scope of review in this case.” The goods subject to review in this case were therefore limited to those for which the original decision maintained registration and in respect of which the applicant sought review.
This treatment reflects the rule limiting the scope of review in non-use cancellation proceedings. A non-use cancellation review does not automatically entail a comprehensive re-examination of the original cancellation decision. The scope of review depends on the relief sought by the party applying for review and the portion of the decision with which that party is dissatisfied. Where the applicant challenges the portion maintaining registration, the review concerns the goods for which registration was maintained; where the registrant challenges the portion cancelling registration, it must separately file an application for review in accordance with law. A portion not placed under review will ordinarily not automatically be re-examined in review proceedings initiated by the other party.
This procedural rule has practical significance for both registrants and applicants. For a registrant, where the original non-use cancellation decision has cancelled registration for certain goods and the registrant considers that cancellation improper, it need to timely seek review and may not expect the matter to be resolved in review proceedings brought by the applicant against the portion maintaining registration. For an applicant, the grounds for review and evidentiary challenges need to focus on the goods for which the original decision maintained registration, with specific arguments directed to the correspondence and similarity between those goods and the goods actually used.
V. Strategy for Challenging Evidence: The Relationship Between Procedural Objections and Substantive Rebuttals One of the applicant’s principal grounds for review was that, “after extensive searches, it found no actual use of the mark under review,” and that it had not been afforded an opportunity to examine and challenge the evidence of use submitted by the respondent; it therefore sought review and requested such an opportunity. This ground has procedural significance, but the key issue in a non-use cancellation review remains the substantive determination. An opportunity to examine and challenge evidence may assist the applicant in scrutinizing and rebutting the evidence, but it cannot substitute for a specific evidentiary analysis.
In a non-use cancellation review, an applicant seeking effectively to rebut a registrant’s evidence of use need to raise specific arguments directed to the critical links in the evidentiary chain. First, it needs to examine whether the evidence falls within the prescribed period. The prescribed period in this case was from October 29, 2021 to October 28, 2024; materials outside that period can serve only a corroborative function. Second, it needs to examine whether the evidence displays the mark under review itself, or whether the sign used is substantially identical to the mark under review in terms of identification. Third, it needs to examine whether the user is attributable to the registrant, a licensee, or another party whose use is not contrary to the registrant’s will. Fourth, it need to examine whether the goods actually reflected in the evidence are the same as or similar to the goods subject to review. Finally, it needs to examine whether the relevant use constitutes public use in commercial activities and is capable of performing a source-identifying function.
In this case, CNIPA ultimately accepted the respondent’s evidence, indicating that the applicant’s evidentiary challenges failed to undermine the overall probative value of the evidence of use with respect to time, user, sign, goods, and source-identifying function. The case demonstrates that an evidentiary challenge in a non-use cancellation review need to not stop at the procedural assertion that the evidence “was not subject to examination and challenge,” but need to address specific weaknesses in the evidentiary chain. Where the registrant’s evidentiary system already forms a relatively complete closed loop, the applicant must present more refined substantive rebuttals, particularly as to whether the goods actually used and the goods subject to review constitute similar goods.
Conclusion: Implications for Strategies to Clear Blocking Marks Through Non-Use Cancellation The “Huawei Cloud Yaoguang” case provides a concrete example of a strategy to clear blocking marks following issuance of the Non-Use Cancellation Guidelines. Non-use cancellation remains an important avenue for addressing genuinely idle trademarks. However, in relation to large technology companies, cloud-service brands, software platforms, artificial-intelligence products, or enterprise technology signs, applicants must assess the likelihood of use with greater caution. Ordinary search engines, e-commerce platforms, and corporate-status searches may form the basis of an initial investigation, but they may not be sufficient to cover the actual-use scenarios of technology trademarks.
Applicants for non-use cancellation need to expand the scope of their searches in accordance with industry characteristics. For trademarks relating to software, cloud computing, and artificial intelligence, searches need to not be confined to product sales pages, but need to also cover the registrant’s official website, product center, developer documentation, technical white papers, cloud marketplaces, industry media, technical blogs, developer communities, official press releases, awards and honors, partner pages, conferences and events, tendering and bidding information, and solution materials. For trademarks with pronounced technological or B2B characteristics, evidence of use may be found in commercial promotion, technical communications, and industry applications, and may not take the form of transaction evidence typical of ordinary consumer goods.
The case also demonstrates that an assessment of the expected benefit of a non-use cancellation application need to take account of the rule maintaining registration for similar goods. Even where an applicant determines that a trademark is used only for one specific technology product, it need to continue to analyze whether that use may cover other similar items among the designated goods. Where the goods actually used are similar to a large number of designated goods, even a successful non-use cancellation application may result only in cancellation for some goods. An applicant with a clear objective of removing a blocking mark need to, before filing, assess whether the target goods may be protected by the spillover effect of the registrant’s evidence of actual use for similar goods.
More broadly, a non-use cancellation strategy needs to be coordinated with invalidation, opposition, litigation, and administrative referral mechanisms. Where the principal defect of a blocking trademark is that it is idle and unused, non-use cancellation remains directly applicable. Where the blocking trademark also involves free-riding on another brand, conflict with prior rights, or bad-faith registration, reliance on non-use cancellation alone may not address the central defect in the case. Strategies for clearing blocking marks after 2025 place greater emphasis on an applicant’s integrated assessment of trademark status, likelihood of use, similarity of goods, and procedural objectives.