AI Companies Gain Ground in Copyright Battles Amid Ongoing Legal Struggles
By Blake Montgomery, The Guardian, July 1, 2025
Last week marked a significant turn in the ongoing legal battles surrounding artificial intelligence and copyright law, as several leading tech companies secured favorable rulings in court. These victories, particularly for AI firms using copyrighted text for training their models, highlight the complexity and evolving nature of intellectual property rights in the age of AI.
Anthropic and Meta Score Court Wins
In a notable case, a U.S. judge ruled that Anthropic, the developer behind the Claude chatbot, did not violate copyright laws by using books to train its AI system without obtaining authors’ permissions. Judge William Alsup likened Anthropic’s use of these books to “a reader aspiring to be a writer,” framing the training process as analogous to a human learning from existing works rather than outright infringement.
Shortly thereafter, Meta received a similar boost from the judiciary. U.S. District Judge Vince Chhabria, presiding in San Francisco, determined that the plaintiffs had failed to convincingly demonstrate that Meta’s AI would cause "market dilution" by flooding the market with derivative works. This ruling significantly weakens copyright holders’ claims that AI-generated content could directly undermine their original works’ value.
On the same day as Meta’s ruling, a group of writers initiated a lawsuit against Microsoft, accusing the company’s Megatron text generator of copyright infringement. However, given the precedents set by the Anthropic and Meta cases, these plaintiffs face uphill challenges in proving their claims.
Broader Legal Battles Across Media Types
These cases represent skirmishes within a broader, more complex legal war over AI’s use of copyrighted materials spanning different media. Disney and NBCUniversal recently sued Midjourney over its AI image and video generation tools, alleging unauthorized use of iconic characters such as Darth Vader and the Simpsons. Similarly, major music labels — Sony, Universal, and Warner — are taking legal action against Suno and Udio, companies that produce AI-generated music.
Meanwhile, the New York Times continues its lawsuit against OpenAI and Microsoft regarding AI-generated textual content. Legal experts note that while text-based AI tools are beginning to see favorable rulings for tech companies, image and audio-related cases may pose more difficulties because AI models allegedly replicate copyrighted imagery more directly.
John Strand, a trademark and copyright lawyer at Wolf Greenfield, explained, “The specific media involved — whether written works, images, videos, or audio — significantly impacts the fair-use analysis in each case. For example, the market for books is very different from that for movies, and this difference is becoming a key factor in court decisions.”
Anthropic’s Controversial Training Methods
The court documents also revealed a startling detail about Anthropic’s data acquisition strategies. The company reportedly pirated an estimated seven million books to build its training database. In an unusual remedial step, Anthropic purchased physical copies of these books, scanned them to digitize their content, and then destroyed the hard copies. Critics view this as a literal and destructive manifestation of the immense content consumption required by AI developers.
Ars Technica noted that while more meticulous and less destructive digitization approaches exist, they are slower and less aligned with the AI industry’s imperative to “move fast and break things.”
Additional Technology and Legal Updates
Aside from the AI copyright battles, last week saw noteworthy developments in other tech-related legal issues:
-
Trump’s Gold Phone Rebranding: The promotional website for the "T1" gold phone, associated with Donald Trump, dropped its previous “Made in America” claim. The branding now reads “proudly American,” echoing Apple’s subtle “Designed in California” tagline. This shift reflects the complicated logistics and supply chains involved in smartphone manufacturing, which rarely meet strict “Made in USA” standards.
-
U.S. Supreme Court Upholds Online Age-Verification Laws: The Supreme Court ruled in favor of Texas’s online age-check law, which requires adults to verify their age before accessing sexually explicit content. The decision, by a 6-3 majority, supports nearly two dozen states’ similar regulations. Justice Clarence Thomas emphasized the state’s interest in protecting children from exposure to adult material. However, critics warn this ruling could lead to increased restrictions on online sexual health information and LGBTQ+ content, raising concerns over internet freedom and censorship.
Looking Ahead
As courts continue to define the boundaries between AI innovation and copyright protections, many questions remain about how legal standards will evolve across different types of media. The fast-moving pace of AI development, combined with the intricate legal frameworks governing intellectual property, suggests that these battles are far from over.
Experts will be closely watching forthcoming cases, including the ongoing lawsuits involving AI-generated imagery and music, to gauge how the legal landscape adapts to new technologies. Meanwhile, the industry’s appetite for large-scale data consumption and rapid product development will continue to raise ethical and legal challenges that demand careful scrutiny.
For further updates on technology and law, subscribe to The Guardian’s TechScape newsletter.
