News

True or false? China issues draft guidance on use of references and citations for advertisements

offices employment
offices employment

On 12 December 2025, China's State Administration for Market Regulation (“SAMR”) issued draft guidance on the use of referenced content in advertisements (the “Draft”). The Draft aims to further standardize the use and presentation of references in advertising activities and to enhance the operability of relevant provisions of China's Advertising Law—particularly Article 4, prohibiting false or misleading advertisements and Article 11(2), requiring data, statistical information, survey results, abstracts, quotations, and other referenced content to be truthful, accurate, and accompanied by clearly identified sources, as well as applicable scopes or validity periods where relevant. The public consultation period recently closed on 12 January 2026, and we expect the final version to be released later this year, with only minor changes expected.

Introduction

The Draft contains 22 articles and sets out detailed rules across several core areas. These provisions are built around the fundamental principle that any cited content must be truthful, accurate, and lawful; advertisers must not use fabricated, falsified, or unverifiable references, nor may they indicate citations in a way that deceives or misleads consumers.

Key provisions

The key provisions under the Draft are as follows:

1) Clearer standards for different types of citations

The Draft sets out detailed standards governing how different types of referenced content may be used, including requirements for the source of the content and the methods by which it is obtained. For example:

  1. Data obtained from experiments: the Draft provides requirements on the institute providing the data, the measuring instruments, facilities and environments, as well as the method used for testing or experimentation..
  2.  Statistical information and survey results: These must be collected through scientific methods, with comprehensive and reasonable scope. Where sampling is used, the sample must meet requirements of scientific rigor, representativeness, and universality, and any deviation must remain within reasonable limits.
  3. Abstracts and quotations: These must be reproduced verbatim. The referenced materials must be authentic, verifiable, and consistent with established scientific knowledge.

2) Clearer rules for  indicating sources and limitations

The Draft provides practical guidance on how advertisers must indicate the source of different types of citations, such as experiments, institutions, academic articles, magazines, or online postings, and clarifies when limitations on scope or validity periods must be expressly stated.

3) Requirements for clear and prominent presentation

The way cited content is expressed (whether in text, images, audio, or video) must be clearly recognizable under normal viewing conditions. Limitations concerning product characteristics, applicable scopes, or validity periods must not be displayed in smaller fonts, lighter colours, or any form that reduces visibility.

4) Narrower exemption for use of superlatives

The Draft imposes further controls over the use of superlative claims: advertisers may not rely on overly narrow or artificially defined market segments to factually justify the use of extreme statements or superlatives, such as “best” or “No. 1” in advertisements. In effect, under the Draft, the safe harbour for superlatives (contained in SAMR's 2023 Guidelines on Enforcement of Absolute Terms) which permits factual, time‑ or territory‑limited claims such as “No. 1 by market share in 2024 in X region” does not apply where the relevant segment, region, or market slice is considered artificially fabricated as too narrow under objective standards and would mislead consumers about the business's true market position or competitive advantage etc.

5) Criteria for determining misleading use of citations

The Draft identifies circumstances in which the use of citations may constitute a false or misleading advertisement. For instance, an advertisement will be deemed misleading if the source of the citation cannot be identified or the referenced content cannot be verified. Likewise, where statistical data or survey results only apply only to specific regions, time periods, industries, or product scopes, failure to disclose such limitations will be a significant factor in determining the advertisement misleading.

6) Prohibition on Citing fraudulent or improper sources

Citing content from news reports, academic papers, research studies, or similar materials that were published or disseminated through paid news arrangements or academic fraud by the advertiser is prohibited. Advertising authorities must refer such cases to the regulators responsible for news, technology, and related industries.

7) Exemption for advertiser‑generated data (including AI-generated data)

Advertisers are generally exempt from indicating the source of data, statistics, or survey results when such content is generated by the advertisers themselves through their own experiments, measurements, statistics, or surveys. Nevertheless, the advertiser remains fully responsible for proving the truth of the claim, providing supporting evidence during enforcement, and avoiding misleading impressions. It is also interesting to note that the data, information, and conclusions generated by using artificial intelligence (AI), deep synthesis, or other similar methods in advertisements are considered self-generated content.

8) Liability for third‑party service providers

Entities engaged by advertisers to conduct experiments, measurements, statistics, surveys, or similar activities will be considered an advertising agent if they knowingly produce data or conclusions intended for advertising use and shall bear the corresponding liability under the Advertising Law..

9) Enforcement discretion for administrative authorities

Importantly, the Draft allows enforcement authorities to exercise discretion in imposing penalties and may choose not to punish the parties for minor violations. For example, cases where administrative authorities may choose to be more lenient could be cases where infringing advertisements were published only briefly with limited exposure and which are promptly corrected by the advertiser, or first‑time violations causing minimal harm that are immediately rectified etc.

Conclusion

Overall, the Draft is highly detailed and operational, reflecting SAMR's intention to strengthen the reliability and transparency of referenced content in advertisements. It directly tackles long‑standing issues such as the misuse of extreme or superlative terms and the use of smaller font sizes or less prominent colours to obscure key limitations. In doing so, the Draft significantly improves the practical enforceability of the relevant Advertising Law provisions and provides clearer expectations and limits for both advertisers and their service providers.

Additionally, from a compliance perspective, businesses should begin assessing how they collect, verify, and present cited data in their advertising materials, and should be prepared to adjust internal review processes once the final rules are issued. Particular attention will be needed in industries where product claims frequently rely on technical data, survey results, comparative statements or superlatives.

We will continue to monitor the release of the final version as well as how the rules are applied in practice, especially given the wide enforcement discretion retained by local authorities across China. Their divergent interpretations and enforcement styles will likely shape how strictly the new citation standards operate in different regions.

If you would like a full English translation of the new Draft or have other questions please do not hesitate to contact our lawyers listed in this article.


AI tools have been used to support research, drafting, and editing of this publication. All content has been reviewed and approved by Hogan Lovells lawyers.

Authored by Grace Guo, Stefaan Meuwissen, and Eugene Low.

View more insights and analysis

Register now to receive personalized content and more!