search("recent AI legal cases outcomes", recency_days=7)

How copyright law could threaten the AI industry in 2024 | Reuters (www.reuters.com)


quote_lines(63, 67)
quote_lines(73, 77)
quote_lines(81, 85)
quote_lines(95, 101)

Jan 2 (Reuters) - If 2023 was the year that artificial intelligence changed everything, 2024 could go down as the year that U.S. copyright law changes AI.

The explosion of generative AI and the popularity of products from Microsoft-backed 【22†(MSFT.O)】 OpenAI, Meta Platforms 【23†(META.O)】, Midjourney and others led to a spate of copyright cases by writers, artists and other copyright holders who say AI has only succeeded thanks to their work.

Judges 【24†so far】 have 【25†been skeptical】 of the plaintiffs' infringement claims based on the content generated by AI. But courts have not yet addressed the trickier, potentially multi-billion-dollar question of whether AI companies are infringing on a massive scale by training their systems with reams of images, writings and other data scraped from the internet.

How copyright law could threaten the AI industry in 2024 | Reuters

Several groups of authors have filed proposed class-action lawsuits this year over the use of their text in AI training. They 【26†include writers】 ranging from John Grisham and "Game of Thrones" author George R.R. Martin to 【27†comedian Sarah Silverman】 and 【28†former Arkansas governor Mike Huckabee】.

Similar lawsuits have also been filed by copyright holders including 【29†visual artists】, 【30†music publishers】, stock-photo provider 【31†Getty Images】 and 【32†the New York Times】.

They all argue that tech companies infringe their copyrights by taking and reproducing their materials without permission for AI training. The plaintiffs are asking for monetary damages and for court orders blocking the misuse of their work.

Tech companies have hired legions of lawyers from some of country's largest law firms to fight the cases. They have defended their AI training in comments to the U.S. Copyright Office, comparing it to how humans learn new concepts and arguing that their use of the material qualifies as "fair use" under copyright law.

"Just as a child learns language (words, grammar, syntax, sentence structure) by hearing everyday speech, bedtime stories, songs on the radio, and so on, a model 'learns' language by being exposed — through training — to massive amounts of text," Meta told the office.

AI proponents also argued that adverse rulings would be disastrous for the industry, which they say has relied on a reasonable assumption that copyright law protects their data handling.

An ongoing lawsuit involving Thomson Reuters 【33†(TRI.TO)】 — the parent company of Reuters News — could be one of the first major bellwethers for AI copyright issues.

The information-services company accused Ross Intelligence in 2020 of illegally copying thousands of "headnotes" from Thomson Reuters' Westlaw legal research platform, which summarize points of law in court opinions, to train an AI-based legal search engine.

A federal judge 【34†ruled in September】 that the Delaware case must go to trial to determine whether Ross broke the law. The case could set a key early precedent on fair use and other questions for AI copyright litigation.

A jury could begin to hear the case as early as next August.

AI lawsuits explained: Who's getting sued? (www.techtarget.com)


quote_lines(107, 115)
quote_lines(135, 163)
quote_lines(171, 175)

Recently, a bevy of legal action demanding compensation from AI companies has been filed in the U.S. and Europe. The plaintiffs include authors, artists and major media organizations who have consistently 【52†expressed concern about AI stealing their work】 and producing mediocre derivatives.

An 【53†open letter from the Authors Guild】 -- signed by more than 8,500 authors, including Margaret Atwood, Dan Brown and Jodi Picoult -- urges tech companies responsible for 【54†generative AI】 applications, such as ChatGPT and Bard, to cease using their works without proper authorization or compensation. The authors want companies to pay for the data they scraped for training -- the "food" for AI systems, endless meals for which there has been no bill.

Authors also express concern that generative AI threatens their profession by 【55†flooding the market with machine-written content】 based on their work. This was a problem in recent months as Amazon took action against AI authors 【56†spamming†www.vice.com】 the bestseller list with generated works.

Prior to the release of the Authors Guild letter, two North American authors -- Mona Awad and Paul Tremblay -- filed a lawsuit against OpenAI, claiming the organization 【57†breached copyright law】. The suit argued that OpenAI breached copyright law because ChatGPT generated accurate summaries of the author's works and, therefore, must've trained on the authors' works. They aren't the only ones. Author and comedian Sarah Silverman is also suing OpenAI and Meta for illegally reproducing her memoir, The Bedwetter, without permission. But that argument may not hold up in court because of the way generative AI works.

Individual authors and artists aren’t the only plaintiffs. In December 2023, The New York Times became the first major American news publication to sue OpenAI for using copyrighted works in AI development.

AI lawsuits explained: Who's getting sued?

Numerous cases have been brought against generative AI companies regarding copyright and misuse. Here are some of the companies being sued.

### GitHub, Microsoft and OpenAI

A class-action suit was filed against these companies involving 【65†GitHub's Copilot】 tool. The tool predictively generates code based on what the programmer has already written. The plaintiffs allege that Copilot copies and republishes code from GitHub without abiding by the 【66†requirements of GitHub's open source license】, such as failing to provide attribution. The complaint also includes claims related to GitHub's mishandling of personal data and information, as well as claims of fraud. The complaint was filed in November 2022. Microsoft and GitHub have repeatedly tried to get the case dismissed.

### Stability AI, Midjourney and DeviantArt

A complaint against these 【67†AI image generator】 providers was filed in January 2023. The plaintiffs alleged the systems directly infringe on plaintiffs' copyrights by training on works created by the plaintiffs and creating unauthorized derivative works. The complaint also takes issue with the fact that the tools can be used to generate work in the style of artists. The judge on the case, William Orrick, said he was inclined to dismiss the lawsuit.

### Stability AI

In January 2023, Getty Images issued a 【68†complaint†copyrightlately.com】 against Stability AI for allegedly copying and processing millions of images and associated 【69†metadata】 owned by Getty in the U.K. Getty filed another lawsuit against Stability AI in the U.S. District Court for the District of Delaware days later, which raised many copyright- and trademark-related claims, and pointed to "bizarre or grotesque" generated images that contained the Getty Images watermark and, therefore, damaged Getty's reputation.

### OpenAI

Authors Paul Tremblay and Mona Awad are suing OpenAI for allegedly 【70†infringing on authors' copyrights】. Butterick is one of the attorneys representing the authors. The complaint estimated that more than 300,000 books were copied in OpenAI's training data. The suit seeks an unspecified amount of money. The case was filed in June 2023.

### OpenAI and Microsoft

The New York Times is suing OpenAI for copyright infringement. The case, filed December 2023, alleges that millions of New York Times articles were used to train and develop OpenAI’s chatbot and other technology, which now competes with the news organization as a source of reliable information. The case also alleges that OpenAI’s language models mimic the Time’s style and recites its content verbatim. The Times is the first major American news outlet to sue OpenAI and Microsoft for copyright infringement. The Times approached the companies earlier in the year to discuss the copyright issue but never reached an agreement.

### Meta and OpenAI

Sarah Silverman's lawsuit against Meta and OpenAI alleged copyright infringement and said ChatGPT and 【71†Large Language Model Meta AI】 (Llama) were trained on illegally acquired data sets with her work contained. The suit alleges the books were acquired from shadow libraries, such as Library Genesis, Z-Library and Bibliotek, where the books can be torrented. Torrenting is a common method of downloading files without proper legal permission. Specifically, Meta's language model, Llama, was trained on a data set called the Pile, which uses data from Bibliotek, according to a 【72†paper†arxiv.org】 from EleutherAI, the company that assembled the Pile. The suit was filed in July 2023.

### Google

A class-action lawsuit is being brought against Google for alleged misuse of personal information and copyright infringement. Some of the data specified in the lawsuit includes photos from dating websites, Spotify playlists, TikTok videos and books used to train Bard. The lawsuit, filed in July 2023, said Google could owe at least $5 billion. The plaintiffs have elected to remain anonymous.

  * Does training a model on copyrighted material require a license? Generative AI systems make copies of the training materials as part of the training process. Does that interim copying require a license, or is it fair use?
  * Does generative AI output infringe on copyright for the materials on which the model was trained? If generative output constitutes a derivative work or infringes the training data's reproduction right, then it infringes on copyright. Courts will need to rule whether similarities in output and training data are derived from protected materials or unprotected materials. Who is liable for copyright infringement when AI infringes?
  * Does generative AI violate restrictions on removing, altering or falsifying copyright management information? The 【73†Digital Millenium Copyright Act】 provides restrictions on removal or alteration of copyright management information, such as watermarks. This is exemplified in the Stability AI case, where the watermark reproduced by Stable Diffusion on generated works constituted false copyright management information.
  * Does generating work in the style of someone violate that person's rights? This is known as the right of publicity, which varies from state to state. It prohibits the use of someone's likeness, name, image, voice or signature for commercial gain.
  * How do open source licenses apply to training AI models and distributing the resulting output? The plaintiffs in the Copilot case argued that republishing Copilot training materials without attribution -- and not making Copilot itself open source -- violates 【74†open source license】 terms.

Legal Battles Over AI Copyright Challenge Tech Giants and News Industry | Cryptopolitan (www.cryptopolitan.com)


quote_lines(185, 199)
quote_lines(205, 211)

  * AI-generated content sparks copyright lawsuits as creators claim their work is unlawfully used to train AI systems, leading to uncertain legal outcomes.
  * The New York Times sues OpenAI and Microsoft, alleging their AI systems produce content similar to Times articles without permission, intensifying copyright concerns.
  * Data access has become a key competitive factor in the AI industry, with companies like Adobe and Microsoft well-positioned to secure data licensing agreements.

In a rapidly evolving technological landscape, artificial intelligence (AI) and【76† copyright law intersection】 has become a focal point of debate and legal contention. Recent lawsuits have questioned whether AI systems capable of generating human-like prose infringe on intellectual property rights. This article delves into the legal challenges surrounding AI-generated content, the implications for the industry, and the potential consequences for the businesses involved.

Over the past year,【77† authors,】 a prominent photo agency, and now major news organizations have taken legal action, asserting that their intellectual property has been illicitly used to train AI systems. These AI systems, which can produce text resembling human writing and fuel applications such as chatbots, have become the subject of court scrutiny.

## The New York Times takes legal action 

The New York Times recently filed a lawsuit on December 27, alleging copyright infringement against OpenAI and Microsoft. This marks the first significant challenge by a major American news outlet regarding the use of AI in this manner. The lawsuit contends that OpenAI’s ChatGPT and Microsoft’s Bing Chat have the capability to generate content nearly indistinguishable from articles published by The Times. This raises concerns that these tech giants are exploiting The Times’ substantial investments in journalism without obtaining permission or making payments.

At the heart of these legal battles lies the “fair use” doctrine, a fundamental aspect of intellectual property law. The doctrine allows creators to build upon copyrighted works as long as they make substantial transformations and do not enter the same market as a substitute for the original creator. However, the application of this doctrine to AI tools remains largely uncharted legal territory.

Legal experts acknowledge that there is no clear-cut answer as to whether AI systems’ use of copyrighted material constitutes infringement or fair use. This lack of legal clarity has led to numerous lawsuits with potentially billions of dollars at stake. While these legal proceedings are still in their early stages, a final resolution may be years away, especially if appeals lead to varying decisions among circuit courts.

Legal Battles Over AI Copyright Challenge Tech Giants and News Industry | Cryptopolitan

The outcome of these lawsuits carries substantial implications for the burgeoning AI industry. Some AI companies have seen a surge in venture capital investment, with the public rollout of AI language models like ChatGPT drawing significant attention. For instance, OpenAI is reportedly considering a stock plan to value the company at over $80 billion. Microsoft has invested $13 billion in OpenAI and incorporated its technology into its products. However, concerns about using intellectual property to train AI models have been top of mind for investors.

## Data ownership and competitive advantage 

Access to substantial datasets has become a pivotal factor in the competitive AI landscape. Companies like Adobe, Bloomberg, Meta, and Google, which own extensive data or have accumulated datasets, are now developing AI tools. Established companies like Microsoft are well-equipped to secure data licensing agreements and navigate legal challenges. In contrast, startups with limited capital may struggle to obtain the data required to compete effectively.

The resolution of these lawsuits may significantly shape the future of copyright law and AI development. Whether these legal battles serve as catalysts for changes in copyright regulations or primarily function as negotiation tactics remains to be seen. The legal landscape’s evolution will play a crucial role in determining the direction of the nascent yet heavily capitalized AI industry.

AI Copyright Legal Outcomes