In a significant move to tighten its grip on artificial intelligence, China's internet regulator has removed over 14,000 AI products during the initial phase of a robust cleanup initiative known as "Qinglang," which translates to "Clear and Bright." This campaign represents a dramatic step towards increased regulatory actions concerning AI technologies within the country.

Understanding the Importance of Qinglang

The Qinglang campaign is an annual effort led by the Cyberspace Administration of China (CAC) aimed at eliminating harmful or illegal online content. This year's focus marks a historical moment, as it targets AI for the first time on such a large scale.

During this initial phase, the CAC accomplished a comprehensive sweep, which included:

  • Removal of over 14,000 non-compliant AI products, including websites, applications, and AI agents
  • Eradication of more than 6 million instances of illegal or harmful information
  • Suspension of 26,000 accounts and deactivation of 1,300 AI product listings
  • Elimination of nine illegal open-source datasets

The crackdown initiated in April 2026 aims to address four critical areas: failure to register AI models, insufficient safety filters, AI data poisoning, and inadequately labeled AI-generated content.

Tech Giants Adapt to New Regulations

Major technology companies in China have rapidly adapted to the new regulations. For instance, Huawei has incorporated thorough reviews in its app store, while Alibaba has enhanced its content identification protocols. Zhipu developed a new review model to meet compliance standards, and DeepSeek implemented checks to prevent data manipulation.

In contrast, ByteDance's platforms, like Doubao, chose to disable specific features instead of complying with new requirements for anti-addiction measures. Local regulatory bodies have also adjusted their strategies, with Beijing implementing self-checks supplemented by routine monitoring and Shanghai adapting rules based on individual platform needs.

What Lies Ahead in the Qinglang Campaign

The second phase of Qinglang is expected to escalate the regulatory approach further. It will focus on addressing AI's role in the dissemination of misinformation, production of violent content, impersonation, harm to minors, and execution of paid astroturfing campaigns. Stricter penalties for non-compliance have been promised by the CAC.

A new set of rules will be enforced starting July 15, which includes measures surrounding AI applications. As these changes unfold, stakeholders in the tech industry will need to stay vigilant about compliance and regulatory expectations moving forward.

Disclaimer: This material is for informational purposes only and does not constitute financial advice.