Anthropic, an AI startup founded in 2021, has secured a significant US$1.5 billion ($2.28 billion) settlement in a Class Action copyright lawsuit. The legal action was launched in 2024 by novelist Andrea Burtz and with non-fiction writers Charles Graeber and Kirk Wallace Johnson.
If the settlement receives approval from the judge, the company will distribute approximately USD 3,000 to the author for each of the estimated 500,000 books contained in the contract.
Additionally, we will destroy illegally downloaded books and refrain from adopting pirated titles in future chatbot training.
The settlement is the largest copyright agreement in US history and sets a key legal benchmark for the evolving dynamics between AI Enterprises and its content creators.
The impact will undoubtedly extend to a variety of other ongoing copyright lawsuits against AI companies, including Openai, Microsoft, Google, and more recently Apple. In June, Meta successfully defended in a copyright dispute, but the ruling was opened for a potential litigation.
The settlement follows a groundbreaking ruling issued in June 2025 on AI development and copyright, which portrayed the distinction between legal AI training and illegal acquisition of content. The allegations against humanity suggest that more than 7 million books were illegally downloaded from two online Shadow Libraries between June 2021 and July 2022.
Officials are expected to finalise the list of authors whose works will be compensated by September 15th.
Cautious optimism
In Australia, responses to potential reconciliation have been tempered by careful optimism. Stuart Glover, policy director for the Australian Association of Publishers, commented:
While we welcome these court enforcement measures against accountability, the settlement underscores the need for AI companies to respect copyright and to compensate creators fairly.
Furthermore, for Australian authors and publishers whose works were exploited without remuneration, this represents the persuasive appeal of the Australian government to ensure that AI companies remain copyrighted and that they will reward them for their use.
Lucy Hayward, CEO of the Australian Authors Association, added:
Although the full details remain private, the settlement is a promising recognition that AI companies simply cannot work for authors and artists to develop broad linguistic models.
Lucy Hayward emphasizes the need for ongoing compensation for Australian authors whose activities are being used to train AI models.
Australian Authors Association
Nevertheless, it is important to note that authors will only be compensated if the publisher registers their work with the US Copyright Office within the specified time frame. Hayward has expressed concern about the provision. This is written by many of the seven million works pirated by international authors, saying that “many global authors suspect that they could forfeit settlement compensation.”
She is urging the Australian government to enact new laws that require tech companies to “provide ongoing compensation for creators of Australian textbooks used in offshore model training.”
Legal risks
In June, US Judge William Allsup ruled that adopting books for AI training was not a violation of US copyright law. However, he concluded that humanity still faces trials over the use of pirated materials to curate its library.
Since then, Judge Alsup has criticized the settlement, citing its shortcomings. He said he will schedule a subsequent hearing on September 25th, “I will grab my nose and see if I can approve it.”
Failure to secure settlement approval can expose humanity to more severe financial debt. The exam is scheduled for December. If the company loses, US copyright law allows statutory damages of up to $150,000 per infringement for a willful violation.
Wolters Kluwer legal analyst William Long predicts that the potential damages in the trial could escalate to billions, posing the risk of putting the company at risk or bankruptcy.
Recently, humanity has secured a new funding round worth US$13 billion, pushing its valuation to $183 billion. Keith Kupferschmid, president and CEO of Copyright Alliance, argues that this “authentic companies can afford to compensate copyright holders without compromising their ability to innovation and compete.”
For Mary Raysenberger, CEO of the Author Guild, the groundbreaking settlement represents “a great victory for the author, publisher and rights holder.” Rasenberger predicts that “paves the way for more licenses to grant both rewards and autonomy about how settlements work by AI companies, as they fit into a properly functioning free market.”
Take a step forward
This particular settlement could provide slight support to Australian authors and publishers whose works are not registered with the US Copyright Office, but it is undoubtedly a potentially lucrative news for global creators. This means progress towards establishing a structured licensing framework.
Australia's copyright law has no exceptions to US-style “fair use.” AI companies claim to protect their training methods. Calls for legislative adjustments have arisen as major AI players, including Google and Microsoft, are pushing copyright immunity on the Australian government.
A recent interim report by the Productivity Committee proposed exceptions to text and data mining in Australian copyright law, allowing AI training on protected Australian works. This proposal encountered considerable opposition from the Australian Authors Association and the publishing department.
As Minister of Arts Tony Burke said in August 2025, the government has copyright protections “without plans, no intentions, no desire to undermine.”
The Australian publishing industry is not entirely opposed to AI development, but important legal and ethical challenges continue. The Australian Publishers Association advocates government policies on AI that prioritizes the parents of creators, aiming for a balanced approach that enables both AI innovation and the cultural industry to flourish, with advocates for clear ethical frameworks, transparency, appropriate incentives, and prioritizes creator parents.
Agata Mrva-Montoya is the president of the Roundtable on Information Access for People with Printing Disabilities.
This article was originally published in conversation. Please read the original article.
Source link: inkl.com.
