References
Campbell v. Acuff-Rose Music Inc. 510 U.S. 569 (1994). Justia U.S. Supreme Court Center. https://supreme.justia.com/cases/federal/us/510/569/
This landmark case established that parody can be protected under the fair use doctrine, even if used commercially. The ruling focused on transformative use, where new meaning or expression is added, making it particularly relevant for AI-generated content that transforms existing works.
Cao, P., Zhou, F., Song, Q., & Lu, Y. (2024). *Controllable Generation with Text-to-Image Diffusion Models: A Survey.* arXiv. https://doi.org/10.48550/arxiv.2403.04279
This survey reviews advancements in text-to-image diffusion models, explaining how they generate unique images by refining random noise. The emphasis on the originality of outputs, even when trained on existing datasets, supports the argument that AI-generated works can be creative and transformative.
Cooper, A. K., & Coetzee, S. (2020). *On the ethics of using publicly-available data.* In O. Branco et al. (Eds.), *Responsible design, implementation, and use of information and communication technology* (pp. 159-171). Springer. https://doi.org/10.1007/978-3030-45002-1_14
This chapter discusses the ethical concerns of using publicly available data, particularly in balancing accessibility with privacy. It explores the risks and responsibilities of using such data in AI contexts.
Feist Publications Inc. v. Rural Telephone Service Co. Inc. 499 U.S. 340 (1991). Justia U.S. Supreme Court Center. https://supreme.justia.com/cases/federal/us/499/340/
This case clarified that while facts cannot be copyrighted, compilations of facts can be protected if they show originality. This is critical for AI, as it highlights the boundaries of copyright protection in data-driven contexts.
Goyal, M., & Mahmoud, Q. H. (2024). *A Systematic Review of Synthetic Data Generation Techniques Using Generative AI.* *Electronics,* 13(17), 3509. https://doi.org/10.3390/electronics13173509
This review examines various methods of synthetic data generation, a growing area in AI that helps bypass copyright concerns. By mimicking statistical patterns of real data without duplication, synthetic data allows AI to generate unique content while protecting privacy and intellectual property.
Hatamizadeh, A., Song, J., Liu, G., Kautz, J., & Vahdat, A. (2024). *DiffiT: Diffusion Vision Transformers for Image Generation.* arXiv. https://doi.org/10.48550/arxiv.2312.02139
This paper introduces DiffiT, a diffusion-based AI model for generating high-quality, original images. The model’s iterative process emphasizes the creation of new visual content, highlighting how AI can generate innovative works rather than reproducing existing data.
Klosek, K., & Blumenthal, M. S. (2024, January 23). *Training generative AI models on copyrighted works is fair use.* Association of Research Libraries (ARL). https://www.arl.org/blog/training-generative-ai-models-on-copyrighted-works-is-fair-use/
This article argues that using copyrighted works to train generative AI models falls under the fair use doctrine, particularly due to the transformative nature of the outputs. It also discusses how AI can advance technology and innovation without harming the market for the original works.
Loving, T. (2023, March 30). *Current AI copyright cases – Part 1.* Copyright Alliance. https://copyrightalliance.org/current-ai-copyright-cases-part-1/
This article outlines ongoing lawsuits against companies like Stability AI, which use copyrighted works to train generative AI models. These cases are pivotal in determining how AI’s use of copyrighted material will be treated under current and future copyright laws.
Navaroli, A. C. (2024, April 23). *AI’s most pressing ethics problem.* Columbia Journalism Review. https://www.cjr.org/tow_center/op-ed-ais-most-pressing-ethics-problem.php
This op-ed discusses the ethical concerns of using synthetic data to train AI models, particularly how it may reinforce biases. It highlights the potential dangers of using AI-generated data as input for further training, calling for regulation to mitigate these risks.
Sony Corp. of America v. Universal City Studios Inc. 464 U.S. 417 (1984). Justia U.S. Supreme Court Center. https://supreme.justia.com/cases/federal/us/464/417/
This case ruled that providing technology capable of both infringing and non-infringing uses (such as VCRs) does not automatically result in liability. This ruling is relevant to AI models, which can be used for both legitimate and potentially infringing purposes, emphasizing that the technology itself is not inherently illegal.
Media Attribution
Figure 1: OpenAI ChatGPT Screen Recording Recording of Chat Interaction with ChatGPT of OpenAI https://openai.com/chatgpt/
Figure 2: Collidu Ethical Decision Making Presentation Ethical Decision Making [PowerPoint Slides]. Collidu. https://www.collidu.com/presentation-ethical-decision-making
Figures 3, 4, 6 - 12, 17, 19, 22: Original Images by OpenAI's DALL-E, Modified by Edmond Leaveck Images generated using OpenAI's DALL-E with modifications made by Edmond Leaveck. Personal attributions.
Figure 5: Screen Capture of OpenAI Session Screen capture of OpenAI chat session. This image was created as an example and does not reflect biases. Personal screen capture.
Figure 13: Microsoft Bing Search Suggestions Screen Capture Screen capture of Microsoft Bing search suggestions through Microsoft Edge browser. Personal screen capture.
Figure 14: NVIDIA Generative AI Glossary Generative AI Glossary. NVIDIA. https://www.nvidia.com/en-us/glossary/generative-ai/
Figure 15: Calls9 AI History Timeline The history of AI: A timeline from 1940 to 2023. Calls9. https://www.calls9.com/blogs/the-history-of-ai-a-timeline-from-1940-to-2023
Figure 16: Brooks, J. L., Groening, M., & Simon, S. (Producers). (1994). *Treehouse of Horror V* (Season 6, Episode 6) [TV series episode]. In D. Silverman (Director), *The Simpsons*. Gracie Films; 20th Century Fox Television. Available on Disney+.
Figure 18: Business2Community Counterfeit Brands Article Counterfeit brands from China and in China. Business2Community. https://www.business2community.com/branding/counterfeit-brands-from-china-and-in-china-046056
Figure 20: Stanford University State of AI Charts The state of AI: 10 charts. Stanford HAI. https://hai.stanford.edu/news/state-ai-10-charts
Figure 21: Gretel AI Synthetic Data Glossary What is synthetic data? Gretel AI. https://gretel.ai/technical-glossary/what-is-synthetic-data
Figure 23: OpenAI Sora Demo Video Released as part of marketing materials for OpenAI's Sora.
Figure 24: 1X Tech Androids: NEO Androids: NEO. 1X Tech. https://www.1x.tech/androids/neo