What China’s pursuit of fairer AI means for marketers

A new set of regulations on AI recommendation algorithms went into effect in China on March 1 in an attempt to introduce unprecedented oversight and inject transparency and accountability into an opaque industry. This is the first case of a major economy enforcing such sweeping rules on the machine, and the world is watching.

China’s new AI regulation is part of an ambitious government-led tech reset that started in late 2020, which saw the introduction of laws in data privacy, data security, antitrust, and more. The latest round of AI regulation marks another monumental step toward building the type of squeaky-clean internet that Chinese regulators envision.

Poorly designed AI could lead to negative consequences.

Without broader oversight, algorithms could knowingly, or unknowingly, damage not only individual consumer experience but also social well-being at large.

The Chinese government has been concerned about algorithmic bias—systematic errors that create unfair outcomes privileging one group of users over another. Tech companies have allegedly charged consumers different prices for the same item or for ride-hailing services.

  • AI designers, without a full grasp of the issues at hand and the variables needed to be considered, could end up building algorithmic models making inadequate decisions.
  • The blind pursuit for profitability could result in AI programmers writing codes that prioritize sales at the expense of consumers’ physical or financial well-being.
  • Using a biased dataset to train AI can also lead to problematic results or biased outcomes caused by human subjectivity.

But AI is here to stay.

  • From augmenting customer experiences through AI-powered avatars to improving customer service efficiency, automation has no doubt brought many benefits to businesses and society.
  • The introduction of the new laws governing AI recommendations reinforces the importance of AI in the future of China’s economy and acknowledges that greater governance is needed to actualize the technology’s full potential.

Here are a few key takeaways for tech companies and marketers from the new regulation:

  • Don’t encourage overindulgence. The new regulation bans algorithms that encourage excessive consumption.
  • Protect consumer rights. The new regulation intends to end pricing discrimination—the practice of selling products or services to different individuals at a unique price based on the premise of their data profile or shopping history. It also bars discriminatory and biased user tags.
  • Be transparent with the AI’s intent. Regulators are requiring platforms to provide their users disclosure of purposes and motives of their recommendation algorithms while providing users the option to delete user tags or opt out of it altogether.
  • Adhere to previous approaches to tech governance. The measures echo ethos trumpeted by watchdogs in recent years. Other provisions dissuade algorithms that cause harm to vulnerable segments of the population such as minors and seniors, promote monopolistic business practices, induce internet addiction, promote unsavory content, and spread news from unlicensed sources.

Can China set the standard for the regulation of AI?

While China’s AI governance model is unlikely to be cut and pasted into another country due to the differences in local politics and business environments, it is a grand undertaking worthy of observation and providing some inspiration.

However, it is uncertain whether there will ever be a perfectly unbiased AI, as the system is always going to be vulnerable to some degree of human subjectivity. The November 2021 launch of the Shanghai Data Exchange promises greater access to data by creating a pilot for a data marketplace for tech companies, but an era of unbiased datasets may be more of a goal than an achievable reality.

Platforms and marketers in China will have to take a long-term view and find creative approaches to AI. At times, they will also have to endure short-term pain for long-term gain. For example, since October, TikTok’s Chinese sister app Douyin has featured mandatory 5-second pauses that urge users to take a break after long sessions, per South China Morning Post.

Of course, issues will arise, loopholes will surface, and amendments will come. For now, regulators in China have made their priority clear: market fairness, consumer protection, and national security.

The full regulation is available in Chinese via official WeChat post or in English as translated by Standard University.

"Behind the Numbers" Podcast