The Ethical Implications of AI on Creative Professionals
Expanding on my current studies in Ethical Computing Technologies in Society, I decided it would be a good idea to compile my research and talks into what I hope will be a useful discussion. As an IT professional with many years of experience in the industry, as well as someone who grew up in a family of successful writers, artists, and musicians, I find this topic especially important to me as a professional who plays (at least a small) role in the adoption and development of AI technologies today. It is reasonable to assume that all IT professionals have a role in the integration of AI tools in business and personal life, and I feel that talks like these are critical as our IT landscapes expand and change for the better or worse.
OpenAI’s ChatGPT and DALL-E have transformed AI into digital content and art creation, blurring the line between human and machine innovation. This has raised concerns about AI models plagiarizing existing artists’ work and replacing their skills with AI models that can produce artwork quickly. Nightshade, a tool that protects artists’ intellectual property rights by “poisoning” AI models like Midjourney, GPT, and DALL-E, has emerged as a disruptive tool in the face of an overabundance of AI models attempting to duplicate human originality and artistic ability. However, the use of AI for creative output may stifle the development of unique and inventive ideas, potentially jeopardizing future opportunities for painters, writers, content creators, and other artists in various lucrative industries (Coeckelbergh, 2023).
The ethical implications of AI models in art creation include concerns about the authenticity and originality of artistic works, as well as the potential harm to AI model authors’ intellectual property. The introduction of Nightshade and similar tools has serious ethical and legal ramifications for the creative industry, as well as job losses and future job opportunities in an evolving AI landscape (Peter, 2023).
Nightshade, an AI model, is used to protect artists’ intellectual property rights by “poisoning” models like Midjourney, GPT, and DALL-E by altering their training data through website scraping. This results in a degradation of their performance and outputs. The poisoned art is then disseminated to the internet, where models systematically scrape for new content consume it as training data. This degrades the AI model’s output, preventing individuals and businesses from effectively adopting these models (Heikkilä, 2023).
Nightshade’s research and development include data poisoning techniques, which introduce unexpected behaviors into machine learning models while training. These samples, visually identical to benign images but containing significant differences in image data, impair the model’s ability to generate meaningful images. For example, if the prompt “Dog” is given, the poisoned model could return a picture of a cat instead.
Artists can now download Nightshade from the University of Chicago’s website to test their digitized artwork before it is published online. The tool, along with Glaze, adds digital “noise” to an image, making it unreadable by AI, so to speak. These tools enable artists to influence AI model training by disrupting the traditional process. This is similar to how AI model creators scraped the web without giving artists control. Artists inject poisoned data onto the web, destabilizing corporations’ work without giving them control (Shan et al., 2023).
Some Real-world Examples & Implications
In 2022, the Colorado State Fair’s annual art competition featured AI-generated artwork, including Jason M. Allen’s “Théâtre D’opéra Spatial.” This was the first time AI-generated artwork had won a prize, prompting outrage among artists who accused Allen of cheating. AI’s ability to mimic established artists’ styles and create customizable artworks using text prompts has raised concerns about originality, validity, ownership, and the future of artist employment opportunities. AI can analyze large amounts of existing artwork, blurring the line between inspiration and reproduction, undermining the importance of human creativity and originality (Roose, 2022).
Jason M. Allen — Colorado State Fair
The New York Times has criticized OpenAI and Microsoft for stealing copyrighted articles to build their AI models, sparking debates about data privacy and ownership. This has resulted in legal disputes between writers, authors, and artists who are fighting back against the systematic appropriation of their work by tech business AI models. While copyright law protects authors’ rights, AI-generated content has the potential to expand artistic possibilities and accessibility. However, recent events highlight the importance of regulating AI technology, as the line between inspiration and imitation blurs, potentially leading to conflicts and issues of intellectual property, copyright ownership, fair use, and the loss of creative job opportunities (De Vynck & Izadi, 2023).
In January 2023, artists sued text-to-image model creators Midjourney and Stable Diffusion for stealing and profiting from their work without permission. The disregard for artists’ rights and livelihoods creates a risky precedent for the creative community and industry. It is critical to investigate artists’ well-being and the consequences of exploitation of their work, as well as the future role of businesses in providing visibility and distribution options (Tangermann, 2023).
The UK Supreme Court ruled in December 2023 that an AI contribution to a patent cannot be considered the invention of the patent, instead allowing only natural persons to be designated as inventors. The debate revolves around whether AI should be granted intellectual property rights, with some arguing that AI should serve society rather than intellectual ownership (Reporter, 2023).
The Ethical Considerations
The ethical implications of AI-generated art include authorship, originality, intellectual property infringement, and job loss. AI-generated art challenges traditional notions of creativity and the value of human artists. It can disrupt traditional creative markets and raise concerns about human labor in the creative process. AI-generated art can also diminish the originality of human creation, violating human authors’ intellectual property rights by using existing works to create new art.
Originality is another concern, as AI-generated art has the potential to replicate the uniqueness and distinction of human artists. Companies like Microsoft and OpenAI have been accused of exploiting artists through their AI art generation technologies, disrupting job opportunities for creatives. This raises ethical concerns about intellectual property rights, fair compensation, and the impact on the artistic community and their work. As AI technologies advance, questions of ownership, responsibility, and accountability arise, necessitating the establishment of transparency, guidelines, regulations, and laws to protect artists, their livelihoods, rights, and creativity.
Artists must strike a balance between using tools for creative advancement while remaining authentic and ethical, and using tools for commercial espionage. The proper application of AI tools requires a thorough examination of how they affect artistic expression, originality, and the broader implications for the creative community.
So, how do we deal with it?
Artists have expressed concerns about the impact of AI art generators on the industry, ethical issues, and the potential devaluation of creative effort. Critics argue that AI-generated art reduces job opportunities, undervalues artists’ work, and lacks originality. Some artists have even called for boycotts to protect artistic integrity and human creativity. Artists are actively advocating for stronger rights for themselves and their work through activism, educational initiatives, and negotiations with AI generator platforms (Shaffi, 2023).
The intersection of AI enterprises and artists has created moral issues due to the complicated relationship between technological innovation, artistic integrity, and ethical considerations. AI art generators present challenges by potentially undervaluing human artists’ work, exploiting labor without fair compensation, and blurring the distinction between genuine artistic creation and AI-generated replicas. Artists navigate the use of AI technology by advocating for transparency, ethical practices, and fair compensation. A balanced approach is required to protect artists’ rights and interests while maximizing AI’s potential for artistic expression (Guild, 2023).
The rise of AI-generated art has profound implications for the artistic community, challenging traditional notions of creativity, originality, and authorship, as well as undermining the ethical implications of individuals’ ability to earn a living while maintaining the integrity of their craft. The shift in artistic practices and impact on artistic identity will result in difficulties in expressing their uniqueness and the worth of their work, resulting in a loss of artistic identity, jobs, authorship, and purpose.
Ethical AI development should prioritize transparency, data protection, and accountability to reduce the possibility of AI-generated art infringing on creatives’ rights while ensuring technology breakthroughs meet ethical standards. Collaboration among creatives, politicians, and AI developers is critical for developing a better understanding of the ethical issues raised by AI-generated art.
References
Abbott, R. (2021, September 10). https://artificialinventor.com/first-patent-granted-to-the-artificial-inventor-project/. Retrieved March 5, 2024, from https://artificialinventor.com/first-patent-granted-to-the-artificial-inventor-project/
Coeckelbergh. (2023). The Work of Art in the Age of AI Image Generation: Aesthetics and Human-Technology Relations as Process and Performance. https://journals.open.tudelft.nl/jhtr/article/view/7025
De Vynck, G., & Izadi, E. (2023, December 28). New York Times Sues OpenAI, Microsoft for Using Articles to Train AI. Washington Post. Retrieved March 5, 2024, from https://www.washingtonpost.com/technology/2023/12/27/new-york-times-sues-openai-chatgpt/
Heikkilä, M. (2023, December 19). https://www.technologyreview.com/2023/10/23/1082189/data-poisoning-artists-fight-generative-ai/. MIT Technology Review. Retrieved March 5, 2024, from https://www.technologyreview.com/2023/10/23/1082189/data-poisoning-artists-fight-generative-ai/
Shan, S., Cryan, J., Wenger, E., Zheng, H., Hanocka, R., & Zhao, B. Y. (2023, February 8). Glaze: Protecting Artists from Style Mimicry by Text-to-Image Models. arXiv.org. https://arxiv.org/abs/2302.04222
Peter. (2023). Generative AI tools in art education: Exploring prompt engineering and iterative processes for enhanced creativity. https://digitalcommons.lindenwood.edu/faculty-research-papers/477/
Reporter, G. S. (2023, December 20). AI Cannot Be Named as Patent ‘Inventor’, UK Supreme Court Rules. the Guardian. Retrieved March 5, 2024, from https://www.theguardian.com/technology/2023/dec/20/ai-cannot-be-named-as-patent-inventor-uk-supreme-court-rules
Roose, K. (2022, September 2). An A.I.-Generated Picture Won an Art Prize. Artists Aren’t Happy. (Published 2022). The New York Times. Retrieved March 5, 2024, from https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html
Tangermann, V. (2023, January 18). Artists Sue Stable Diffusion and Midjourney for Using Their Work to Train AI That Steals Their Jobs. Futurism. Retrieved March 5, 2024, from https://futurism.com/artists-sue-stabile-diffusion-midjourney
Vincent, J. (2018, October 23). https://www.theverge.com/2018/10/23/18013190/ai-art-portrait-auction-christies-belamy-obvious-robbie-barrat-gans. The Verge. Retrieved March 5, 2024, from https://www.theverge.com/2018/10/23/18013190/ai-art-portrait-auction-christies-belamy-obvious-robbie-barrat-gans
Wadhwani, P., & Pathak, A. (1970, January 1). AI Image Generator Market Size, Share Analysis Report 2032. Global Market Insights Inc. Retrieved March 5, 2024, from https://www.gminsights.com/industry-analysis/ai-image-generator-market
Shan, S., Ding, W., Passananti, J., Wu, S., Zheng, H., & Zhao, B. Y. (2023, October 20). Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models. arXiv.org. https://arxiv.org/abs/2310.13828
Thubron, R. (2023, October 25). Nightshade tool can "poison" images to thwart AI training and help protect artists. TechSpot. https://www.techspot.com/news/100618-nightshade-tool-poisons-images-thwart-generative-ais-help.html
Shaffi, S. (2023, January 23). ‘It’s the opposite of art’: why illustrators are furious about AI. The Guardian. https://www.theguardian.com/artanddesign/2023/jan/23/its-the-opposite-of-art-why-illustrators-are-furious-about-ai
Guild, G. A. (2023, September 27). AI Art Generators: Where We Stand — The Graphic Artist Guild. The Graphic Artist Guild. https://graphicartistsguild.org/ai-art-generators-where-we-stand/
