China Targets AI-Generated Slop in Digital Content Sweep

The Cyberspace Administration of China (CAC) has initiated a substantial campaign to curb what it terms "slop" - low grade, mass produced synthetic content - that is increasingly prevalent across social media and news feeds. This action by the country's principal internet watchdog signifies a wider effort to manage the misuse of generative artificial intelligence technology.

Credit: UNSPLASH

A statement released on the CAC's official website detailed the outcomes of its most recent enforcement operation. Over 5.61 million instances of content deemed either harmful or illegal were removed from online platforms. Additionally, the operation resulted in the deletion of approximately 49,000 user accounts.

This regulatory sweep encompassed a substantial number of digital venues, affecting 2,400 websites and various applications. Prominent Chinese digital spaces were targeted, including Douyin, the sister application to TikTok, its rival Kuaishou, RedNote, and Tencent Holdings' widely used WeChat platform. Authorities concentrated their efforts on restricting content that included deepfake news generated by artificial intelligence, overtly explicit and violent material, instances of impersonation, and safety infringements concerning minor users. So these enforcement actions covered a broad spectrum of digital misconduct.

The underlying issue stems from the growing accessibility of generative artificial intelligence. As the tools employed for content creation become more affordable and sophisticated, individuals can quickly generate highly convincing deepfakes and various forms of clickbait. This capability presents significant challenges for established content moderation systems and practices across numerous platforms.

The CAC cited specific examples of content deemed objectionable. Creators operating on platforms such as Douyin, Kuaishou, and WeChat had utilised artificial intelligence to transform respected literary classics, including *Romance of the Three Kingdoms* and *Journey to the West*, into sensationalised, poor quality "digital swill." This content was primarily designed to artificially boost traffic and engagement.

Other problematic accounts were found to have employed AI generated "digital humans" and fabricated news narratives. These materials were subsequently used in deceptive marketing programmes specifically aimed at older internet users, according to the regulator's report. And the scope of the problem extends beyond mere entertainment to include potential financial exploitation.

Chinese internet platform operators have begun to reinforce their internal safety protocols. The CAC noted that developers responsible for leading chatbot technologies, such as ByteDance's Doubao and Alibaba Group Holding's Qwen, have implemented tighter review processes for training data. These developers have also introduced more stringent prohibitions on content output and integrated clearer markers for synthetic materials. Alibaba is the owner of the South China Morning Post.

The Cyberspace Administration of China has committed to ongoing monitoring of the risks associated with content generated by artificial intelligence and synthetic media. Measures will be taken to prevent the misuse and malicious application of artificial intelligence. Simultaneously, the organisation stated its intention to support the continued development of this technology.

• China's Cyberspace Administration initiated a campaign against AI generated "slop" content.

• Over 5.61 million pieces of content and 49,000 accounts were removed.

• The clampdown targeted 2,400 websites and apps, including Douyin, Kuaishou, RedNote, and WeChat.

• Focus areas included AI generated fake news, explicit content, impersonation, and minor safety violations.

• Internet platform operators are strengthening internal guard rails and content identification for AI generated material.

Source: SCMP