By Artists

1. Andersen v. Stability AI, Midjourney, Runway, DeviantArt filed Jan 13, 2023 in ND California over the use of art (via LAION-5B and LAION-400M) to train diffusion models (Stable Diffusion, Midjourney).
2. Zhang v. Google filed Apr 26, 2024 in ND California over the use of art (via LAION-400M) to train a diffusion model (Imagen).

By Writers

1. Tremblay v. OpenAI filed Jun 28, 2023 in ND California over the use of books (via Books3) to train an LLM (ChatGPT).
2. Kadrey v. Meta filed Jul 7, 2023 in ND California over the use of books (via Books3) to train an LLM (Llama).
3. Silverman v. OpenAI filed Jul 7, 2023 in ND California over the use of books (via Books3) to train an LLM (ChatGPT). Consolidated with Tremblay v. OpenAI.
4. Chabon v. OpenAI filed Sep 8, 2023 in ND California over the use of books (via Books3) to train an LLM (ChatGPT). Consolidated with Tremblay v. OpenAI.
5. Chabon v. Meta filed Sep 12, 2023 in ND California over the use of books (via Books3) to train an LLM (Llama). Consolidated with Kadrey v. Meta.
6. Authors Guild v. OpenAI filed Sep 19, 2023 in SD New York over the use of books (via Books3) to train an LLM (ChatGPT).
7. Huckabee v. Meta, Microsoft, Bloomberg, EleutherAI filed Oct 17, 2023 in SD New York over the use of books (via Books3) to train LLMs (Llama, BloombergGPT). Case against EleutherAI was voluntarily dismissed Dec 26, 2023.
8. Alter v. OpenAI, Microsoft filed Nov 21, 2023 in SD New York over the use of books (BooksCorpus) to train an LLM (ChatGPT). Dismissed Nov 7, 2024.
9. Basbanes v. Microsoft, OpenAI filed Jan 5, 2024 in SD New York over the use of nonfiction books (via Books3) to train an LLM (ChatGPT). Consolidated with Authors Guild v. OpenAI.
10. O'Nan v. Databricks filed Mar 8, 2024 in ND California over the use of books (via Books3) to train an LLM (MPT).
11. Nazemian v. NVIDIA filed Mar 8, 2024 in ND California over the use of books (via Books3) to train an LLM (NeMo Megatron–GPT).
12. Dubus v. NVIDIA filed May 2, 2024 in ND California over the use of books (via Books3) to train an LLM (NeMo Megatron–GPT).
13. Bartz v. Anthropic filed Aug 19, 2024 in ND California over the use of books (via Books3) to train an LLM (Claude).
14. Farnsworth v. Meta filed Oct 1, 2024 in ND California over the use of books (via Books3) to train an LLM (Llama).

By Actors and YouTubers

1. Millette v. Nvidia filed Aug 14, 2024 in ND California over the use of videos (from YouTube) to train an AI product (Cosmos).
2. Vacker v. ElevenLabs filed Aug 29, 2024 in Delaware over the use of actors' voices to train a voice-generating model.

By Publishers

1. Getty Images v. Stability AI filed Feb 3, 2023 in Delaware over the use of images to train a diffusion model (Stable Diffusion).
2. New York Times v. Microsoft filed Dec 27, 2023 in SD New York over the use of news articles to train an LLM (ChatGPT). Complaint shows memorization of plaintiff's work.
3. The Intercept v. OpenAI filed Feb 28, 2024 in SD New York over the use of news articles (via WebText, WebText2, Common Crawl) to train an LLM (ChatGPT).
4. Raw Story Media v. OpenAI filed Feb 28, 2024 in SD New York over the use of news articles (via WebText, WebText2, Common Crawl) to train an LLM (ChatGPT).
5. Daily News v. Microsoft and OpenAI filed Apr 30, 2024 in SD New York over the use of news articles (from New York Daily News, Chicago Tribune, Orlando Sentinel, San Jose Mercury News, and others, via WebText and WebText2) to train LLMs (ChatGPT and Copilot).
6. Center for Investigative Reporting v. OpenAI filed Jun 27, 2024 in SD New York over the use of news articles (via WebText and WebText2) to train an LLM (ChatGPT).
7. Dow Jones, New York Post v. Perplexity AI filed Oct 21, 2024 in SD New York over unauthorized scraping of news websites after refusal to license content.

By Music Companies

1. Universal Music Group v. Anthropic filed Oct 18, 2023 in ND California over the use of song lyrics to train an LLM (Claude). Complaint shows memorization of plaintiff's work.
2. RIAA v. Suno and Udio filed Jun 24, 2024 in SD New York over the use of songs to train audio-generating models.
3. GEMA v. OpenAI filed Nov 13, 2024 in Regional Court of Munich over the use of song lyrics to train an LLM (ChatGPT).

By Programmers

1. Doe v. GitHub, Microsoft, OpenAI filed Nov 3, 2022 in ND California over the removal of required attribution from code used for LLM (ChatGPT) training, unfair competition, privacy violations. Claims against OpenAI were voluntarily dismissed in July 2024.

---

Other Relevant Cases

These are recent and ongoing cases about piracy, web scraping, and whether generated artwork can be copyrighted—topics relevant to the AI training lawsuits.

• Thomson Reuters v. Ross Intelligence filed May 6, 2020 in Delaware. Thomson Reuters claims Ross scraped a paywalled legal database to train an LLM that helps with legal research. Similar to the above cases except that there is an open question about whether Thomson Reuters's content is copyrightable.
• Thaler v. Perlmutter filed Jun 2, 2022 in Washington DC. Stephen Thaler claims the U.S. Copyright Office should grant copyright for an image he created with AI in November 2018. USCO says copyright is limited to "original intellectual conceptions of the author."
• Meta v. Bright Data filed Jan 6, 2023 in ND California. Meta claimed Bright scraped Facebook and sold the data. The court ruled for Bright Data in Jan 2024 and Meta dropped the case.
• X Corp. v. Bright Data filed Jul 26, 2023 in ND California. Twitter/X claimed Bright scraped and sold its data. The court dismissed X's complaint on May 9, 2024.
• Cengage Learning v. Does 1-50 filed Sep 14, 2023 in SD New York. Cengage claims the maintainers of Library Genesis have been distributing pirated e-books.
• Cengage Learning v. Google filed Jun 5, 2024 in SD New York. Cengage claims Google has been advertising pirated e-books while rejecting ads for legitimate ones.
• Allen v. Perlmutter filed Sep 26, 2024 in Washington DC. Jason Allen claims the U.S. Copyright Office should grant copyright for an image he created with Midjourney by refining a prompt 624 times.