Skip to content
Submit incident
AII-1000-52269
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.

AI-generated plagiarism floods Chinese online literature platforms

Chinese online literature platforms including Qidian and Tomato Novel reported widespread flooding of AI-generated or AI-assisted novels from 2025 onward, with some authors producing dozens or hundreds of works per month. Platforms began removing these works from recommendation lists in early 2026, and prominent writer Zhang Wei warned that AI-enabled plagiarism through content reconstruction at massive scale was damaging the industry's creative ecosystem.

August 2025 · Qidian Chinese Web · Intellectual property and appropriation
August 2025
AII-1000-522693 sources ↓
Quick facts
Category
Intellectual property and appropriation
Country
China
Source
中华网科技公司
Date
August 2025
Drafted by
Raidu Incident Detection

What happened

From August 2025 onward, large numbers of AI-generated or AI-assisted novels flooded Chinese online literature platforms including Qidian, Tomato Novel, and Jinjang Literature City. According to author accounts, writers used models like DeepSeek and Doupack to generate content, with some able to produce dozens or over 100 novels per month by heavily relying on AI for plot generation and text composition.

On August 18, 2026, Qidian removed numerous works showing heavy AI creation marks from recommendation lists and bestseller charts, though works remained available and authors could continue publishing. In February 2026, Tomato Novel disclosed that some accounts engaged in large-scale AI batch creation, producing low-quality and homogeneous content at scale, with accounts publishing over 100 works per day.

Reported impact

Readers paying for chapters reported feeling deceived when discovering content was AI-generated. One author using AI saw his daily reader count drop from 500–600 to 80–90, failing to meet platform bonus requirements. Writers said AI-generated text lacked emotional resonance and human quality, with flat language, overuse of punctuation marks, and inconsistent characterization in longer works.

Network writer and Beijing Writers Association Vice Chair Zhang Wei (pseudonym Tang Jia San Shao) stated that AI plagiarism had become a serious problem in online literature, with AI technology enabling authors to deconstruct bestsellers, refill their frameworks, and generate seemingly new articles within 48 hours at a scale of 5 million characters, describing the flood of low-quality content as "digital swill" polluting information ecosystems.

Index analysis

One author employing AI-assisted writing produced multiple novels simultaneously across accounts, reportedly earning 70,000 to over 100,000 yuan monthly, while authors relying heavily on AI for efficiency found themselves trapped in lower quality output and declining reader engagement as they became dependent on the tool.

Platform editors expressed concern that authors dependent on AI lost motivation to improve writing skill and seemed addicted to false efficiency gains. Platforms including Qidian and Jinjang implemented content policies beginning in 2025 that positioned AI as an assistive tool for research and polish only, forbidding AI use for core narrative generation.

Governance controls

In April 2025, Qidian issued a public notice opposing use of AI or automation to replace human creation of core content, with enforcement measures including recommendation suspension, delisting from charts, content blocking, and removal. In the same month, 16 leading platforms including Qidian, Tomato Novel, and Jinjang released a self-discipline pledge requiring establishment of AI-assisted creation use norms.

Tomato Novel implemented daily character limits for low-level accounts to prevent bulk posting of new works. Yuewen Group's Miaobi tool was designed for research, inspiration-finding, and scene polish only, with reported weekly invocations in the tens of thousands.

Classification

Category
Intellectual property and appropriation
Industry
Media, social platforms and marketing
Country
China
Harm status
Actual harm
Intent
Unknown
Timing
Unknown
Harm types
Economic or property, Reputational
Harmed
Business
Failure modes
Hallucination
Consequences
Content removal
AI systems
DeepSeek, Doupack, Yuewen Miaobi

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

中华网科技公司
唐家三少谈AI洗稿 创新精神受挑战
9 September 2026
凤凰网(凤凰新媒体)
唐家三少:网络文学成AI洗稿重灾区,两天能生成500万字
9 September 2026
凤凰网(凤凰新媒体)
规模大到失控",唐家三少急了
10 September 2026