The State Administration for Market Regulation published the Interim Provisions on Anti-Unfair Competition in Networks on 11 May 2024, effective 1 September 2024. The rules place fake orders and reviews, false publicity, traffic hijacking, platform rules and data rights into specific online scenarios, and AI-generated marketing is covered. For AI companies that acquire customers through search, recommendation and content placement, the text needs to be read before a product goes live.
From conduct types to specific scenarios
Traditional unfair-competition provisions are organised by type of conduct. The network rules push conduct into search, ranking, recommendation, live streaming and social settings. Traffic hijacking, inserted links and forced redirects are identified as recognisable improper methods; fake orders, invented reviews and organised sham transactions target the review system itself. Companies should check their own acquisition path item by item rather than asking broadly whether there is false publicity. The same claim may be assessed differently on different channels.
AI-generated content is no defence
Using a generative model to mass-produce product descriptions, reviews or user comments does not change the character of the conduct. If the content is false or misleading, it remains within the scope of false publicity. Automated generation can make the conduct look more serious because of scale and concealment. AIGC in marketing material should come with human review and factual verification, and prompts, generation logs and review traces should be retained. Fact-checking before placement is the cheapest form of risk control.
Platform rules and data rights
The rules constrain platform operators that use rules, algorithms or technical means to exclude or restrict competition, and they also reach improper means of obtaining and using data. An AI company that also operates a platform should examine whether ranking rules, traffic allocation and data-interface policies have legitimate reasons and transparency. A company operating inside another platform should focus on remedies when traffic is limited, content is copied or data is scraped. Publication of rules and appeal mechanisms are usually the first evidence in a dispute.
Evidence and internal controls
Online unfair-competition cases depend heavily on electronic evidence. Access logs, recommendation records, campaign back-end data and content review records are both defence material and a source of exposure. Companies should build a traceable chain from generation to review, placement and takedown, and bind third-party agencies on contract for fake traffic and managed accounts. The scope and purpose of data returned by channel partners should also be written into agreements, so that compliance responsibility does not sit with the company while control sits elsewhere.
Administrative and civil tracks run in parallel
Online unfair competition can trigger administrative penalties and civil claims by a harmed party. Platform complaints, temporary takedowns and injunction applications often affect business faster than an administrative decision. A company must prepare both for inspection and for proving the legitimacy of its conduct in civil proceedings. The two tracks ask different questions: the administrative process focuses on whether conduct is unlawful, while civil proceedings also examine loss and causation. Campaign strategy, test records and user feedback can serve both.
For marketing-material and competition review, see our data compliance service and AI infringement service; responsibility allocation with channels and agencies can be handled through commercial contracts.
This article is based on the Interim Provisions on Anti-Unfair Competition in Networks published by the State Administration for Market Regulation on 11 May 2024 and effective 1 September 2024, and on public information. It is for reference only and does not constitute legal advice.