Zhang Wei, vice chairman of the Beijing Writers Association, recently published a statement highlighting two major challenges that artificial intelligence (AI) poses to the online literature sector: frequent "AI-assisted plagiarism" and the proliferation of "digital sludge."

His remarks have sparked widespread discussion across the industry. Interviews conducted by Capitalnews with multiple online novelists revealed strong consensus with Zhang's perspective. These authors collectively stated that while they do not oppose AI serving as a tool, they firmly believe AI must never be allowed to replace human creativity. The value of original, hand-crafted writing should be given greater prominence and protection.

AI-Assisted Plagiarism: Generating 5 Million Words in 48 Hours

AI-assisted plagiarism has evolved far beyond simple copy-and-paste operations. The current method involves deconstructing the framework of successful works and using AI to fill in new content, thereby creating what appears to be an entirely fresh article while essentially reproducing the original structure. Zhang Wei emphasized that this process can generate an staggering five million words within just 48 hours. He characterized this practice as a systematic suppression of innovative spirit, stating it is not "creation" but rather "theft," and not "prosperity" but "counterfeiting."

Zhang Wei stressed that such practices damage not only the industry's ecosystem but also undermine the innovative foundation upon which national cultural soft power relies. The term "digital sludge" refers to the massive volume of cheap, nutritionally empty digital content being produced. He likened these AI-generated materials to an invisible corrosive agent systematically polluting the internet's information ecosystem. While they may appear to be merely low-quality content, he warned they represent a gradual erosion of national information sovereignty.

Zhang Wei has called for strengthened industry self-regulation to establish a robust first line of defense, improved regulations and standards to clarify legal boundaries, and focused efforts to protect young readers by severing the pollution chain. He specifically proposed that AI-generated content should be clearly labeled to help readers and platforms distinguish between authentic human creation and assembly-line output.

Authors Voice Concerns: AI Must Not Replace Personal Creativity

Liu Wei, vice chairman of the Shanghai Writers Association, has been engaged in online literature creation since 2003. He and his fellow authors have discussed the devastating impact of AI-assisted plagiarism, describing how carefully constructed plots, well-developed characters, and thoughtfully built worlds can be replicated into thousands of similar works within seconds using just a few commands. This flood of AI-generated content overwhelms various rankings and makes it nearly impossible for original works to gain visibility among readers.

"This phenomenon of bad money driving out good will quickly cause readers to experience aesthetic fatigue with similar plots, characters, and world settings, ultimately losing interest entirely," Liu Wei explained. He believes that allowing AI-assisted plagiarism to continue unchecked constitutes a rapid strangulation of healthy development within original online literature.

The consequences would be severe: authors would either leave the field entirely or abandon innovation, resorting to copying from existing works. When all creative output begins to resemble a single face, reader attrition becomes inevitable.

AI as a Tool: Industry Consensus Emerges

As stated by fellow author Fei Jia, vice president of the Shanghai Online Literature Writers Association, "AI brings convenience, but it also brings mixed results." Throughout this process, every creator should maintain a strong sense of agency and build their own first line of defense—they may fully utilize AI to serve them, but must never allow it to replace their own creative process.

All interviewed authors share this unified perspective: they hold no opposition to AI when used as a tool. For Liu Wei, research involving ancient surnames and tribal origins benefits significantly from AI assistance. Traditionally, consulting reference books and search engines proved time-consuming, labor-intensive, and disruptive to writing flow. However, AI tools can efficiently locate and organize required information from vast data repositories, providing tremendous assistance to writers.

Novelist Qin Lv incorporates tools like Doubao during her creative process to gather foundational materials and supplement obscure knowledge points. She acknowledges that these tools significantly reduce research time. Nevertheless, she has noticed apparent limitations in AI-retrieved content—information compiled from internet sources contains a mixture of authentic and false data, with some historical facts presenting errors, omissions, or inaccuracies. She emphasized that creative work must still prioritize formal literature and authoritative classics as core references rather than relying entirely on AI.

Qin Lv particularly endorses Zhang Wei's proposal for clear labeling of AI-generated content. She views AI tools as supplementary assistance that cannot replace human creativity and eagerly anticipates platforms recommending more content infused with "human touch."

Literary Creation: Rooted in Human Experience

"Literary creation should be human-centered rather than a product of industrialized assembly, borrowing, stitching together, or outright plagiarism," Liu Wei stated. He described literary creation as a voice emanating from the depths of the soul, born from intense emotional fermentation through life's dramatic experiences. Every character, word, and sentence should reflect an author's spiritual journey along life's long road—having witnessed all manner of trials and tribulations while absorbing the full spectrum of joy and sorrow.

Liu Wei urged platforms to do more than merely remove offending content when addressing AI-assisted plagiarism. Simply taking down such works carries little deterrent effect for AI studios with astonishing production capacity. He called for enhanced penalties through contractual provisions that strongly constrain AI-plagiarism behaviors—striking hard at first appearance and making enforcement genuinely painful. Copyright protection regulations must also evolve alongside technological developments and be improved without delay.

Additionally, Liu Wei proposed that platforms strengthen professional ethics training and evaluation systems for authors, establish comprehensive oversight mechanisms, and provide more learning, training, and exchange opportunities. This approach would help writers instinctively reject the lifeless approach of industrialized assembly-line production. Rather than relying purely on algorithmic recommendations, platforms should maintain artificial recommendation channels under specific rules, ensuring recommendation pools offer greater diversity and contain more "human touch."

Fei Jia believes platforms have stronger incentives than individual authors to combat AI-assisted plagiarism because when readers reach their breaking point with low-quality AI-generated content, they will abandon the platform altogether.

Platforms Respond with Targeted Governance Measures

According to Capitalnews research, multiple online literature platforms launched dedicated governance initiatives several months ago. In February, Fanqie Novel published a specialized notice addressing AI-generated mass-produced filler content, explicitly defining the misuse of AI for large-scale batch creation and content assembly as a violation. The notice revealed that certain accounts were abusing AI technology for massive batch production, generating substantial quantities of crude, homogenous, and unreadable low-quality content. Some accounts updated hundreds of works in a single day, severely crowding out display space for quality original content.

In April, Qidian Chinese Network released a notice to its author community, clearly stating opposition to practices that use AI or other automated methods to replace human creators for core content production, disrupt normal creative order, or interfere with standard operations. The platform further strengthened its regulatory management of non-human automated text generation.

Industry insiders indicate that multiple platforms have already removed substantial numbers of AI-generated filler content, and governance over AI-assisted plagiarism continues to intensify industry-wide.

Preserving the "Handcrafted" Creative Spirit

Tang Qiao, associate researcher and deputy director of the Online Literature Research Office at the Chinese Academy of Social Sciences' Institute of Literature, observed: "It is precisely in this era when technology is readily accessible and computing power is omnipresent that handcrafted creation deserves greater attention and recognition."

However, ensuring that handcrafted works receive not only respect but also sustainable viability requires platforms to implement more substantial adjustments in recommendation mechanisms, revenue distribution, and evaluation systems. Tang explained that the fundamental logic of algorithms favors selecting and reusing what works best, continuously replicating market-validated popular patterns and formulas. When increasing numbers of creators choose to rely on algorithmic templates for batch production, those willing to patiently engage with reality and carefully refine their handcrafted creations naturally become rare commodities.

She advocated that while technology may serve as an auxiliary means to improve efficiency and address shortcomings, the primary creative concepts, spiritual core, aesthetic expression, and value positions of works must always remain under human direction and control.

Commentary: Preventing Digital Pollution from Contaminating Creative Sources

The "digital sludge" topic continues to attract sustained attention. Well-known online authors have spoken out against AI's batch generation of "digital sludge," describing these outputs as invisible corrosive agents systematically polluting the internet's information ecosystem.

Digital sludge refers to low-quality content characterized by illogical reasoning and hollow values in the intelligent era. With each iteration of AI tools bringing enhanced capabilities, today's inferior content has become increasingly difficult to identify, and related plagiarism grows harder to detect. Numerous counterfeit versions of classic works circulate widely online—someone dismantles the framework of popular works, feeds it to AI, and generates millions of words of "entirely new old articles" within 48 hours.

Superficial examination reveals vocabulary and sentence structures that do not constitute direct copying, yet closer inspection shows nothing more than parroting. Growing numbers of people recognize that the proliferation of digital sludge not only squeezes out quality content and pollutes the online environment but also affects public cognition. As digital natives, young people surrounded by inferior information face significant challenges in developing deep thinking capabilities, cultivating innovative capacity, and maintaining fine-grained perception of life.

Although digital sludge currently primarily affects leisure reading materials, the "grab-what-you-can" mentality depletes public resources. Regardless of how verbose AI-generated articles may be, they ultimately assemble themselves from humanity's textual treasure trove—a supply that depends on individual creators' intellectual accumulation. When their years of dedicated work drown in the information flood, the market sends a dangerous signal: serious writing doesn't pay, while plagiarism and copying prove more efficient.

Over time, fewer people will be willing to endure the solitary grind of serious writing, and the raw materials available for AI to harvest will correspondingly shrink and degrade—a terrifying process of reverse selection.

Strategies for Combating Digital Sludge

How can digital sludge be withstood? Readers can theoretically vote with their feet, but the high costs of navigating algorithmic recommendations and information cocoons make discrimination prohibitively expensive. Authors can voice their rights, but individual power remains too dispersed against batch accounts, matrix operations, and disguised plagiarism. Online literature platforms must take more meaningful action as the distributors of algorithms, traffic, and revenue. Their orientation directly shapes outcomes.

Recently, major online literature platforms have conducted related governance measures, including suspending recommendations, removing works from rankings, blocking content, and taking down offending pieces. Beyond these steps, questions requiring clear answers abound: How to enhance identification capabilities for inferior content? Should stronger support for original quality content be implemented? How should repeat offenders who produce digital sludge be punished?

Combating digital sludge does not mean rejecting AI. AI can serve as a pen, a tool, or an assistant—but it must never become a parasite that consumes the living waters at the source. Only when original creators receive proper respect can the creative wellspring remain forever flowing.