Introduction
Artificial Intelligence, is the branch of computer science that simulates human behaviour through data collection, processing and algorithms. The main goal is to enable machines to process things like humans.
With the development of technology, various AI functions appear in front of our eyes, like voice assistants or intelligent medical care, this type of AI is perfectly integrated into our lives and helps tremendously. Of course, AI also brings a lot of problems.As a platform painter, I feel quite deeply mainly from AI painting, since the concept began to rise a variety of AI paintings began to appear, which also brings us original painters, not a small dilemma.Imagine finishing a painting after hours or even days of constant changes, and an AI effortlessly generates the same style of work in seconds after scanning and learning, all without your consent or even knowledge.As absurd as it may seem, this is exactly the reality that many artists encounter on the platform.
AI is silently plundering the fruits of original artists’ labour, algorithms are devouring human creations at a rate of millions of times per second, and the “creativity” we are so proud of is being reduced to data fodder for feeding machines. Behind this behaviour is not only the misuse of technology but also the concentration of data, privacy, algorithmic bias and regulatory lag, which have become issues that cannot be ignored in the current digital society.
Redbook AI Painting: What Does It Depend On?
As a new publishing platform in China, Redbook, with its social and content sharing, e-commerce and personalized services, has been widely acclaimed since its release, and more and more people are using it. Over time, coupled with mutual recommendations between users, more and more artists have joined the platform and started to operate accounts, and I also chose to share my work on Redbook.
If it develops properly Redbook has a good chance of becoming one of the biggest areas of activity for painters. But at the end of July 2024, a painter named “Snow Fish”（雪鱼） reported that the official account of Trik AI on Redbook had used his painting “Kuafu chasing the Sun” for AI training without his knowledge, and officially stated that it was an AI study of Chinese traditional mural rock paintings to generate a work of art. Traditional Chinese mural rock colour study after the generation of works，However, after comparing the paintings, it can be seen that the AI-generated paintings are very similar to the original artist and even have a lot of direct plagiarism, some images with symbolic meanings appear abruptly in the AI works, and some of the patterns that the AI can’t understand are even carried by a mass of unexplained lines.After this incident was exposed and fermented, more painters began to check their works, and it turned out that not a few of them had the same experience.
As Kate Crawford said, AI is more of an extractive industry that relies on resources, cheap labour, and massive amounts of data in addition to technology (Crawford, 2021).Taking AI painting as an example, its first step is to collect a large amount of painting data, which covers painting styles and genres from classical art to modern art, and then further analyze and learn from these works through algorithms in terms of composition, color, style and other features.After collecting a certain amount of data, AI painting can imitate human painters to create, and gradually master painting skills through continuous optimization and iteration, after which it can automatically generate paintings according to instructions, and such generated works generally have a high degree of artistry and completion.
But the question is, does the fact that some of this collected data comes from content posted by users on the platform constitute infringement?We can analyze the case of Redbook to show that users are unaware of the fact that AI is using their work, that this data extraction lacks transparency, and that platforms usually don’t explicitly tell users that their work will be used for AI.It is even directly contained in the user terms and conditions of Redbook that the platform can arbitrarily use the content posted on Redbook by users without their authorization. This leads to the creator’s work being used in the case not claiming their rights, legal regulation lags, and the creator’s copyright and other rights and interests are not institutionalized and protected.
The “blind spot” of algorithmic governance: platforms as “referees” in the digital age
Platforms like Redbook are technological promoters even if they are content collectors in AI governance. After users dug deeper into the agreement, they found that 4.2 on content and information authorization in its user agreement states.
“User grants Redbook Company a free, irrevocable, non-exclusive, non-territorial license to use, including store, use, transmit, reproduce, modify, adapt, compile, publish, display, translate, perform, or create derivative works from User Content (Redbook, 2025).”
In the terms and conditions and mentioned that the user grants the right to use the Redbook, and after that, it is written that no personal information will be transferred to a third party. Numerous users believe that this seems to maintain the interests of the user but does not restrict the self platform of the word trap, ambiguous expression is a hegemony clause. This is reflecting the current regulatory lag, algorithmic governance of several blind spots.
- Users have no immediate way of knowing if and when their work has been used.
- Weak rights defence mechanisms, which prevent users from filing a complaint or claiming damages
- Lack of third-party regulators, and platforms monopolize the right to make and interpret rules.
Platform algorithms are no longer merely tools for filtering information, but reality constructors, and in this lack of intervention, platforms form a new type of governor(Just & Latzer, 2016). Such governance often lacks transparency and accountability, resulting in users being unable to assert their rights.
In an era dominated by digital technology, tech companies are pushing for faster and better functionality from AI, arguing that it can outperform humans because it can produce multiple styles of visually striking imagery faster and on command, but does this mean that AI is superior to humans?Although AI painting is powerful enough to mimic brush strokes and generate compositions, it can’t completely replace human painters because it can’t understand the creator’s mood. Painting is not only an accumulation of techniques, but also a kind of expression of emotions and the touch of the soul, and when both of them reach the extreme, a simple image will be given a profound meaning, and these subjectivity, complexity, and cultural accumulation are exactly the reasons why art can’t be shown by AI. These subjectivity, complexity and cultural accumulation are the core of art that cannot be shown by AI.But in the age of rapid information flow, cultural choice no longer belongs to human beings.Algorithms are replacing the judgment that humans were meant to possess(Andrejevic, 2019). The identity of the creator is being marginalized in silence. So in this context, it is all the more important to defend the value of human artistic creation. Today’s AI painting has become the trend of the future, but its development should never be based on the infringement or even plagiarism of creators.
About the Future
Redbook’s AI incident caused collective opposition from creators, at the time of the incident, I was also looking for other platforms where I could freely share my paintings, but I found that most of the platforms where I shared existed this kind of problem that users could exploit without authorizing the platforms, and in the end, I could only pause my sharing on the platforms for the time being.The rest of the original artists have also issued notices to stop updating their works or transferring them to other secure trading platforms, and many users have spontaneously used the “No AI” icon as their avatars to express their anger at AI infringement, in addition to posting statements on their homepages about not authorizing AI.After the efforts of many users, Trik AI has now been taken down from the App Store，The Redbook user agreement is also inconsistent with the Interim Measures for the Administration of Generative Artificial Intelligence Services issued by China’s State Internet Information Office (SNIO) in conjunction with seven departments, which states that intellectual property rights should be respected and that algorithms, data, and platforms should not be utilized to implement monopolize and unfair competitive practices (Ying Zhu, 2023)，The odds are that the Redbook’s strip will be revised, and this incident gets put to rest for the time being.
As can be seen in the case of Redbook, the power of individuals and small groups cannot effectively change the rules of the platform, and AI still lacks effective and harmonized global regulatory measures to deal with the ever-expanding rights of digital platforms.
So what do we as users really need?
- Transparency protocols and accountability mechanisms
Transparency is not only about making AI’s “black box” operations visible but also about building public trust in AI, especially when it comes to AI training, where users should have the right to know and the right to choose. AI developers should disclose the basic principles and design ideas of their algorithms so that users can understand the general framework and logic of their algorithms. AI developers should disclose the basic principles and design ideas of their algorithms so that users can understand their general framework and logic.
- legal constraints and copyright protection
In the case of Redbook AI, it has been subject to relatively effective legal constraints，However, the actual rights of platforms still far exceed existing legal governance(Flew, 2021, pp. 79–86)，There is still a need for more detailed and effective laws defining AI training, and AI should be held accountable for infringement when it occurs，bringing data use into the realm of public regulation(Flew, 2021, pp. 79–86). Creators should also have the right to indicate that the work is not intended to be used for AI training and obtain copyright protection.
- International frameworks and the concept of people-centered AI
In addition to the legal constraints issued by the State, governance at the international level is also very important. UNESCO released the world’s first normative framework for AI ethics, the Ethical Recommendations on Artificial Intelligence, which suggests that AI must be embedded with mechanisms for transparency and fairness (UNESCO, 2024)，This means that platforms have to guarantee human rights when processing data, especially the labour of the original creator, and cannot be divorced from ethical considerations should be trained and developed with respect for creative labour, realizing that the technology should serve humanity, not replace it.
Overall, governing AI is not a matter of fixing its data collection and algorithms, but requires the construction of cultural, power, and labour relations between humans and AI.
Conclude
It’s not AI itself that we’re resisting, as an original artist I’m well aware that AI can be a trigger for inspiration for us as well, but the question has now risen as to whether or not it’s starting to quietly replace us without the artist’s knowledge.By bypassing licensing and applying a set of opaque data access mechanisms that marginalize labour, platform AI is not only threatening the livelihoods of creators but is also competing for autonomy and voice in the cultural sphere.
But it is not unattainable for AI to coexist with humans as equals, and we want to form a more harmonious relationship of rights and labour with them.As the painters collectively revolted at Redbook, AI should be based on respect for creativity and protection of rights, not unilateral expropriation by platforms. AI must be incorporated into a regulatory and ethical framework that has boundaries rather than silently devouring the labour of the original.
References
Andrejevic, M. (2019). Automated Culture. Routledge EBooks, 44–72. https://doi.org/10.4324/9780429242595-3
Crawford, K. (2021). Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press. https://ebookcentral.proquest.com/lib/usyd/reader.action?docID=6478659&ppg=10
Flew, T. (2021). Regulating Platforms (pp. 79–86). John Wiley & Sons.
Just, N., & Latzer, M. (2016). Governance by algorithms: reality construction by algorithmic selection on the Internet. Media, Culture & Society, 39(2), 238–258. https://journals.sagepub.com/doi/full/10.1177/0163443716643157
Redbook. (2025). Xiaohongshu.com. https://agree.xiaohongshu.com/h5/terms/ZXXY20250119002/-1
Sina Visitor System. (2025). Weibo.com. https://weibo.com/u/1322227362
Statement of non-authorization – Redbook. (2024). Xiaohongshu.com. https://www.xiaohongshu.com/explore/675aa97100000000020165c4?xsec_token=ABloqzjH6wFd24lpOVNTXu7FeXUp9SZ0w7DJxcaGPtIng=&xsec_source=pc_user
UNESCO. (2024). Ethics of artificial intelligence. Www.unesco.org; UNESCO. https://www.unesco.org/en/artificial-intelligence/recommendation-ethics
Ying Zhu. (2023, July 10). Interim Measures for the Administration of Generative Artificial Intelligence Services_State Council Departmental Documents_China.gov.cn. Www.gov.cn. https://www.gov.cn/zhengce/zhengceku/202307/content_6891752.htm
Be the first to comment