Skip to main content

The recently released Administrative Provisions on Algorithm Recommendation of Internet Information Service (“Provisions”) substantially affects technology companies by regulating the use of algorithms. Released by the Cyberspace Administration of China and three other government bodies on 4 January 2022, the Provisions come into effect from 1 March 2022 and regulate the use of algorithm technology to strengthen consumer protection and market practice.

Over the last 5 years, China’s digital economy has developed and scaled rapidly. Technology companies continue to emerge and innovate platforms and services for consumers by utilising big data generated from online activities. Specifically, algorithms are common practices that enable consumers to make choices, aligned to their likes and dislikes. For example, a machine learning algorithm records the clicks and patterns of each consumer on a shopping platform and uses such data to recommend new products to the consumer. However, unethical practices of algorithm technology, such as differential pricing and manipulative push recommendations, can damage the rights and interests of consumers and hinder fair competition.

The Provisions regulates the utilisation of algorithm technology by requiring technology companies to adopt healthy practices. Namely, algorithm technology under the Provisions refers to information provided to users through synthesis algorithm, personalised push recommendations, sort selection, scheduling decisions, and retrieval filters. We summarise the key obligations for companies using algorithm technology (“Service providers”).

  1. Robust management and technical system

Service providers shall establish management and technical system for algorithm technology that includes reviewing scientific and technological ethics, information disclosure and user registration, ensuring data security and personal information protection, preventing telecoms and online fraud, monitoring security assessment, handling emergency responses to security incidents, and publicly disclosing algorithm mechanisms.

  1. Allow users to turn off algorithm technology

Users shall be provided with the option to turn off the algorithm technology and the ability to select or delete the tags related to their personal characteristics.

  1. Protection of minors

Algorithms shall facilitate information beneficial to the physical and mental health of minors. Recommended content that causes unsafe conduct or induces bad habits such as internet addiction is prohibited.

  1. Protection of elderly

Algorithms shall provide safe recommendations to the elderly and consider their needs in areas, such as elderly travel, medical treatment, consumption, and handling administrative affairs.

  1. Unreasonable differential pricing prohibited

Consumers shall have the right to fair pricing and rights. Namely, algorithms are prohibited from rendering unreasonable or illegal forms of differential pricing based on consumer behaviour and previous purchases.

Violations can result in fines between 10,000 RMB and 100, 000 RMB, and criminal responsibility can be pursued where the violation constitutes a crime.

Though the Provisions primarily impact technology companies, fast-moving consumer goods (“FMCG”) and health companies should observe such legal obligations in their marketing strategies and advertisement campaigns. Particularly, companies who rely heavily on social media and key opinion influencers to promote FMCG and health products in the China market. Marketing messages should be aligned with the Provisions to ensure content does not damage the rights and interests of consumers – especially with minors or elderly consumers.

This content appears as a courtesy of Horizons Corporate Advisory, a proud member of the China Collaborative Group (CCG Association). It is informational in nature and does not constitute legal advice or establish an attorney-client relationship between you and its author, publisher or any member of CCG. For more information, please visit www.horizons-advisory.com