Chinese Online Novel Platform Embroiled in Controversy for Collecting AI Tra...

TMTPOST--Tomato Novel, a platform owned by ByteDance, sparked a public outcry for asking novel creators for permission to use their works for AI training.

The protocol in question showed that by signing it, creators would allow the use of their entire or partial works—including titles, summaries, outlines, chapters, characters, personal information, and cover images—for AI research, machine learning, model training, and algorithm development.

One author from Tomato Novel expressed concerns that AI could compete with human-created content, stressing that if AI training uses copyrighted works, it may lead to a loss of copyright for the original text.

Another creator warned that supporting AI-generated content might lead to human work being unfairly labeled as "plagiarism."

An author named Xiao Lu expressed deep concerns over the use of their works in AI training. They compare it to seeing their creations—deemed as their "children"—become part of someone else's body.

The impact of AI on content creation remains a hotly debated topic. The Tomato Novel app currently features several AI tools, including AI-based expansion, rewriting, and custom description features. Writers must agree to terms that grant the platform worldwide, perpetual, irrevocable, and free rights to store, use, publish, modify, and distribute content created with these AI tools.

The platform has faced strong criticism from creators due to its call for signing the protocol. Despite the growing trend of using AI for content training, legal hurdles are unlikely to halt technological progress, according to Li Zonghui, vice president of the Institute of Cyber and Artificial Intelligence Rule of Law at Nanjing University of Aeronautics and Astronautics.

Li acknowledged that while AI tools enrich human civilization and enhance data analysis, Tomato Novel's initial approach lacked proper legal considerations. He noted that merely using terms without a clear profit-sharing mechanism does not ensure fair AI training practices, which should empower rather than undermine human writers.

In response to the backlash, Tomato Novel revised its protocol （which date？）, allowing creators to opt out if they had concerns. The updated protocol emphasizes AI as a tool to aid creators, aiming to improve writing efficiency and enhance reading experiences, such as through AI-assisted character naming and research.

Legal experts advise that AI-generated content should be clearly labeled to avoid public confusion or misidentification. They also pointed out that the copyright status of AI-generated content depends on its originality, uniqueness, and intellectual contribution.

The use of AI tools in online literature is not a new thing, and similar technologies have been adopted across various platforms. For instance, in July 2023, China Literature Group introduced its large model "Yuewen Miaobi," designed to assist writers with services and tools. Other platforms, like "Qimao," have collaborated with Baidu to offer AI-assisted writing features.

Despite these advancements, experts emphasize that AI's role should remain supportive rather than replacing human creativity. Xiao Bai Zhi Zhu, another writer, notes that current AI-generated novels lack the nuanced character development and conversational style of human-written works. While AI may replace lower-tier writers, it is unlikely to impact top writers significantly.

Xue Hong, the president of Shanghai Writers' Association, believes that AI's current influence is minimal on long-form novels with complex plots and unique worldviews. The backlash against Tomato Novel's new agreement highlights the broader concerns within the industry about preserving authors' rights and creativity in the face of AI advancements.

Zhai Wei, a legal expert from East China University of Political Science and Law, argues that platforms must protect authors' rights by obtaining proper authorization before using their works for AI training. They warn against exploiting authors' dependence on platforms and emphasize that fair compensation and voluntary agreements are essential for maintaining a healthy creative environment.

In February, China made a groundbreaking ruling on AI-related copyright infringement, with the Guangzhou Internet Court finding an AI company guilty of infringing on Ultraman works. Similarly, the New York Times has sued OpenAI and Microsoft over non-transparent data collection practices for AI training.