A quick look at lawsuits against major AI companies

Any innovative technology is likely to face litigation, but the period of active development of artificial intelligence has been especially harsh in terms of legal disputes.

As companies like OpenAI, Alphabet, Meta, and others engage in an arms race to become the leader in artificial intelligence, a flurry of lawsuits is emerging. In many cases, content providers are accusing one or more companies of stealing their intellectual property to train their large language models.

Some plaintiffs focus on privacy issues. But others, like Elon Musk, argue that AI companies have abandoned their original missions.

Here's how things stand in this rapidly changing field.

Disclaimer: This is a free translation of the text of the Fast Company publication, written by Chris Morris. Read the original text Here. The text also mentions the company Meta, which is recognized in the Russian Federation as an extremist organization.

Translation by the editorial staff of “Technocracy”. Subscribe to our channel “Voice of Technocracy”, to stay up to date with new materials about the AI ​​products market.

Copyright Lawsuits

There are currently more than 25 lawsuits pending against AI companies. One of them (Thomson Reuters v. ROSS Intelligence) was scheduled to begin on August 23 but is now on hold after the judge adjourned the case a day before the trial was scheduled to begin.

Thomson Reuters vs ROSS Intelligence

In May 2020, Reuters sued ROSS, alleging that the artificial intelligence company used data from Thomson Reuters’ legal research platform, Westlaw, to train its AI system. ROSS says it used the material in good faith. It could be a game-changer for litigation that sets a precedent for others to come. The issues are somewhat different from other copyright lawsuits, but similar enough that the jury’s opinions and verdict could influence how lawyers in other cases argue their case.

Sarah Andersen vs Stability AI

The class action lawsuit is against the AI ​​image generator for allegedly using artists' work without their consent. Andersen, along with Kelly McKernan and Carla Ortiz, filed the lawsuit in January 2023, accusing London-based Stability, DeviantArt, and Midjourney of copyright infringement.

Getty Images vs Stability AI

In February, the stock photography agency accused Stability AI of using more than 12 million photos from its library without a license to train its Stable Diffusion AI image-generation system. Getty also accuses the company of trademark infringement, pointing to AI-generated images that include a Getty watermark that could cause confusion in the marketplace.

Doe1 vs Github

Last May, anonymous software developers alleged in court that OpenAI and Microsoft's GitHub stole source code they wrote to create artificial intelligence tools. However, in July 2024, a California federal judge dismissed their copyright claims, saying they had failed to prove the tools could produce work identical to their own. The allegations of breach of contract and violation of an open-source license are still being contested.

Zhang vs. Google

Similar to Andersen v. Stability, this class action lawsuit sees a group of artists suing Alphabet, alleging that Google's text-to-image AI tool was trained using copyrighted works without permission.

JL vs Google

In June, Alphabet won the case, convincing a San Francisco judge to dismiss a class-action lawsuit alleging the use of personal and copyrighted data to train artificial intelligence systems. Following the ruling, the lawyer who filed the lawsuit filed Zhang v. Google.

Richard Kadrey vs. Meta Platforms

The class action lawsuit is a consolidation of three lawsuits against Meta that accuse the social media company of copyright infringement for allegedly copying their books to train its LLaMa model. The plaintiffs include Sarah Silverman, Kadri, and Ta-Nehisi Coates. A similar lawsuit is pending against OpenAI.

OpenAI ChatGPT Trial

Consolidation of three cases against ChatGPT brought by Sarah Silverman, Paul Tremley, and Michael Chabon. The suit is the same as Kadrey v. Meta: copyright and DMCA infringement related to ChatGPT. A settlement hearing is scheduled for mid-September.

Bartz vs Anthropic

Writers and journalists Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson have filed a lawsuit against Anthropic, alleging that the company used their and thousands of other works to train its chatbot Claude. This is the second case filed against Anthropic, following a similar lawsuit filed by music publishers last year.

Millette vs. OpenAI

The class action lawsuit alleges that OpenAI transcribes millions of YouTube videos without the consent of the authors to train its massive language model. “By transcribing and using these videos in this manner, Defendants profit from Plaintiff’s and other Class Members’ data over and over again,” the complaint says.

Abdi Nazemian v. Nvidia Corp.

Nvidia has not escaped the wrath of authors who claim their copyrighted content was used without permission. This class action lawsuit, filed by three authors, alleges that the chipmaker used 196,640 books to train its NeMo AI platform to simulate natural writing styles. NeMo was taken offline last October “due to reports of copyright infringement.”

Dubus vs. Nvidia Corp.

Like Nazemian v. Nvidia, this lawsuit is brought by another author who claims Nvidia's NeMo was trained on their books without permission. On May 29, the case was marked as “related” to Nazemian's lawsuit, increasing the chances that it will become a class action.

O'Nan v. Databricks, MosaicML

Stuart O'Nan, Abdi Nazemian and Brian Keane, who are leading the lawsuit against Nvidia, separately sued MosaicML and Databricks for allegedly unauthorized use of their workbooks to train their large language model. (MosaicML was acquired by Databricks last July.)

Mackay v. Databricks, Inc., MosaicML

In May, fiction authors Rebecca McKie and Jason Reynolds filed a lawsuit almost identical to O'Nan's against Databricks.

Concord Music Group, Inc. et al v Anthropic PBC

Universal Music Corp and seven other major music publishers have sued Anthropic for direct and secondary copyright infringement, alleging that the artificial intelligence company used unauthorized copies of song lyrics to train its chatbot Claude and distributed those lyrics through Claude. A judge is scheduled to rule in October on Anthropic’s motion to dismiss the case and the music publishers’ request for a preliminary injunction that would prevent Claude from providing any data.

Alter vs. Open AI

The three-case consolidation pits several groups of authors against OpenAI and Microsoft. One of the cases was originally filed by the Authors Guild, led by prominent authors John Grisham, George R.R. Martin, and Jodi Picoult. The writers’ professional organization alleged that the companies used their fiction to train ChatGPT. The lawsuit also names as plaintiffs David Baldacci, Michael Connelly, Sylvia Day, Jonathan Franzen, Douglas Preston, and Scott Turow. In February, the case was consolidated with Basbanes v. Microsoft, a lawsuit filed by journalist Nicholas Gage and author Nicholas Basbanes in January.

New York Times vs Microsoft

The news organization claims that millions of its copyrighted materials were used to train massive GPT language models by OpenAI and Microsoft, which run ChatGPT and Copilot, respectively. Legal experts say the case could have a huge impact on the intersection of generative AI and fair use, which could shape how AI models are built moving forward.

Daily News vs Microsoft

On the same day that the New York Times Company filed its lawsuit, the New York Daily News accused Microsoft and OpenAI of using the archive of articles to train its products. The lawsuit was filed on behalf of the Daily News, as well as the Mercury News, Denver Post, Orange County Register, St. Paul Pioneer-Press, Chicago Tribune, Orlando Sentinel, and South Florida Sun Sentinel. In June, OpenAI filed a motion to consolidate the lawsuit with the New York Times case. The judge has not yet ruled.

The Center for Investigative Reporting, Inc. vs OpenAI

The nonprofit news organization filed a complaint that echoes claims by the New York Times and Daily News in June.

Huckabee vs. Bloomberg

Last year, former Arkansas Governor Mike Huckabee (and others) accused Meta, Bloomberg, and Microsoft of training their large language models on their works without permission. The case against Meta and Microsoft was moved to California, leaving Bloomberg as the sole defendant in the lawsuit. In March, the company asked a judge to dismiss the case, arguing that the use of copyrighted works was part of a research project and fell under fair use. The judge has not yet ruled.

Raw Story Media, Inc. vs OpenAI

An alternative news organization filed a lawsuit in February against the developers of ChatGPT for allegedly using thousands of its materials to train the chatbot. A separate but nearly identical lawsuit was filed by the left-leaning news outlet The Intercept.

UMG Recordings vs Uncharted Labs d/b/a/ Udio and UMG Recordings vs Suno

A large group of music labels, including UMG, Capitol Records, Sony Music, Arista, and Warner Music, banded together in June to sue Uncharted Labs, the maker of Udio.com and Suno, a pair of generative AI tools that let people create music based on prompts. The labels claim that the large language models in the tools were trained on their copyrighted music without permission. Separate lawsuits have been filed against each of the tools.

Lerman v. Lovo, Inc.

Voice actors Paul Sky Lerman and Linnea Sage have filed a class action lawsuit against Lovo, accusing the company of selling AI versions of their voices without permission. They claim they were tricked into providing vocal samples to the company and then found their voices being used in YouTube videos and podcasts under false names.

Lawsuits for the protection of personal data

AT vs OpenAI

ChatGPT users are suing the company for allegedly violating state and federal wiretapping laws by collecting data from users who asked ChatGPT questions. The case, filed last September, is similar to another (PM v. OpenAI) that was withdrawn by the plaintiffs last September.

JL vs. Alphabet Inc.

Google's parent company is accused of using millions of internet users' data to train its LLMs, and the case has left observers in awe. In June, the U.S. District Court for the Northern District of California dismissed the case. However, in that decision, Judge Araceli Martinez-Olguin noted that a second, amended (and simplified) complaint would be filed soon. The plaintiffs filed that complaint on July 11, upholding the case.

Elon Musk vs. Sam Altman and OpenAI

Musk dropped the lawsuit against OpenAI and its CEO in June but revived it in August. Musk claims Altman and co-founder Greg Brockman misled him when they created OpenAI and failed to follow through on their promise to open source the company's technology, partnering with Microsoft instead.

“Altman, along with other defendants, intentionally coerced and deceived Musk by exploiting Musk’s humanitarian concerns about the existential danger posed by artificial intelligence,” the lawsuit says. OpenAI has already responded to these claims in a blog post, showing emails that appear to indicate Musk was trying to commercialize OpenAI before he left the company in 2018.

Subscribe to our telegram channel “Voice of Technocracy”Every morning we publish a news digest from the world of IT, and in the evenings we share interesting and useful articles.