Latest news with #CopyrightAlliance
Axios
6 days ago
- Business
- Axios
Behind the Curtain: What does AI owe YOU?
You could easily live without AI. But AI wouldn't exist without you. So does AI owe you for your small part in creating it? Why it matters: This question sits at the very heart of legal, economic, moral and societal debates unfolding before us today — and deep into the future. The answer will unlock everything from court verdicts to the fortunes of financial winners and losers years from now. The big picture: First, the facts. Large language models (LLMs) such as OpenAI's ChatGPT or Anthropic's Claude were built by ingesting the totality of human information online — everything from news stories by Axios to social media postings by you. AI developers call this "training data" because they used the internet — and all its content — to train machines to think, reason and operate like the humans feeding the web. The goal: think better, faster and cheaper than us. Then, unleash these machines everywhere for everything. By 2025, content produced by billions of internet users helped shape the massive digital corpus that modern AI models were trained on — mostly through public, web-accessible data. Think any content — posts, blogs, forums, websites, Wikipedia edits, reviews, and videos. Some content owners — mass producers and individual creators — are suing AI companies for using their content without permission. The litigation — by media companies, music producers, book publishers — is often brought under copyright laws. AI makers defend their use of protected content by invoking long-contested "fair use" principles. Copyright owners have filed close to 50 lawsuits against AI companies in federal court, according to the Copyright Alliance, which represents creators. Some publishers, including News Corp and Axios, have struck deals with OpenAI for use of their content. The Supreme Court could ultimately decide — likely years from now — whether the owners of these LLMs violated federal law with their training tactics. "We're at the top of the first inning of understanding how the courts are going to define fair use as it applies to AI training," Keith Kupferschmid, president and CEO of the Copyright Alliance, tells us. He said cases involving images and music will be more difficult for AI companies to win, while cases involving text are more complex. But what about you? If you ever posted a comment on Reddit, or tweeted, or shared pictures on Instagram, or posted a rant on Facebook, you're a tiny part of the LLMs' brains. Your words were gobbled up to help train the machines, creating untold riches for tech companies and investors. But no one asked your permission or offered to compensate for that tiny part you played in making others rich. It's doubtful you'll get a penny for your time or mind now. President Trump, in his speech at last month's AI summit in Washington, made plain his opposition to companies doling out cash based on derivative use of the internet (unless it's blatant piracy). "[W]hen you read something and when it goes into this vast intelligence machine, we'll call it, you cannot expect to every time, every single time, say: 'Oh, let's pay this one that much,'" he said. "[W]e have to allow AI to use that pool of knowledge without going through the complexity of contract negotiations, of which there would be thousands for every time we use AI." His main point: It would give China an insurmountable edge if we tie ourselves in knots over who owes what to whom. But some AI leaders, most notably Anthropic's Dario Amodei, argue you should benefit — in the future. They believe if AI only enriches the big companies and investors, America will face a rebellion — and potentially sky-high unemployment. So instead of paying you now, which might hamstring the companies, they talk of spreading the wealth if AI truly creates trillions of dollars of wealth long-term. This could be as novel as a guaranteed minimum income (universal basic income, or UBI), funded by AI companies, or as ordinary as higher taxes on AI to fund social programs and government overall. Or it could be a private sector indirect benefit of higher growth, driving higher wages. Regardless, the AI companies will get a lot bigger and a lot richer — before you get a penny, if you ever do.
The Sun
25-06-2025
- Business
- The Sun
US judge: AI training on books is fair use, partly
SAN FRANCISCO (United States): A US federal judge has sided with Anthropic regarding training its artificial intelligence models on copyrighted books without authors' permission, a decision with the potential to set a major legal precedent in AI deployment. District Court Judge William Alsup ruled on Monday that the company's training of its Claude AI models with books bought or pirated was allowed under the 'fair use' doctrine in the US Copyright Act. 'Use of the books at issue to train Claude and its precursors was exceedingly transformative and was a fair use,' Alsup wrote in his decision. 'The technology at issue was among the most transformative many of us will see in our lifetimes,' Alsup added in his 32-page decision, comparing AI training to how humans learn by reading books. Tremendous amounts of data are needed to train large language models powering generative AI. Musicians, book authors, visual artists and news publications have sued various AI companies that used their data without permission or payment. AI companies generally defend their practices by claiming fair use, arguing that training AI on large datasets fundamentally transforms the original content and is necessary for innovation. 'We are pleased that the court recognized that using 'works to train LLMs was transformative,'' an Anthropic spokesperson said in response to an AFP query. The judge's decision is 'consistent with copyright's purpose in enabling creativity and fostering scientific progress,' the spokesperson added. - Blanket protection rejected - The ruling stems from a class-action lawsuit filed by authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson, who accused Anthropic of illegally copying their books to train Claude, the company's AI chatbot that rivals ChatGPT. However, Alsup rejected Anthropic's bid for blanket protection, ruling that the company's practice of downloading millions of pirated books to build a permanent digital library was not justified by fair use protections. Along with downloading books from websites offering pirated works, Anthropic bought copyrighted books, scanned the pages and stored them in digital formats, according to court documents. Anthropic's aim was to amass a library of 'all the books in the world' for training AI models on content as deemed fit, the judge said in his ruling. While training AI models on the pirated content posed no legal violation, downloading pirated copies to build a general-purpose library constituted copyright infringement, the judge ruled, regardless of eventual training use. The case will now proceed to trial on damages related to the pirated library copies, with potential penalties including financial damages. Anthropic said it disagreed with going to trial on this part of the decision and was evaluating its legal options. 'Judge Alsup's decision is a mixed bag,' said Keith Kupferschmid, chief executive of US nonprofit Copyright Alliance. 'In some instances AI companies should be happy with the decision and in other instances copyright owners should be happy.' Valued at $61.5 billion and heavily backed by Amazon, Anthropic was founded in 2021 by former OpenAI executives. The company, known for its Claude chatbot and AI models, positions itself as focused on AI safety and responsible development.
