OpenAI sued for using copyrighted books without permission

From Authors Guild: New documents released last Thursday in Authors Guild v. OpenAI—the major book authors’ ongoing lawsuit alleging OpenAI and Microsoft illegally trained their products on copyrighted books—show that executives and employees at both companies intentionally broke the law and their disdain for human authorship. OpenAI Policy Director Jack Clark stated in May 2020: “Our work on AI and Creativity is going to increasingly lead to us creating systems that substitute for the labor of [] people . . . The better we do on GPT-X, the more worried genre fiction authors will become about us substituting for them on Amazon.” He then doubled down: “Our work in this area will make people unemployed. . . There will be a point where a bunch of artists express worry about what we’re doing here and we’ll likely ignore their concerns and release anyway . . .” https://lnkd.in/gZT9Wiiv #openai #chatgpt #ai #genai #generativeai #authorsguild #copyright

"Stop pretending generative AI is an educational tool. It's not. It never was. This technology harms writers, harms the writers academia has hired to teach writing, and harms students learning to write."

Nothing I didn't already know, but to see the contempt of OpenAI & MS specifically towards genre fiction, it's hard not to take personally. Honestly, it's also hard not to take other teachers' adoption of AI into writing pedagogy personally, either. Is that how little everyone thinks of creative writing as a discipline? That we'd slap "edu" on LLMs and try out this unethical parlor trick on emerging writers, on the future of human storytelling? Willingly use "tools" frankensteined on theft and made by these smug technofascists creeps? Support an industry bolstered by the most ignorant, anti-reading, anti-creativity misogynists of my lifetime? Stop pretending generative AI is an educational tool. It's not. It never was. This technology harms writers, harms the writers academia has hired to teach writing, and harms students learning to write. https://lnkd.in/gTGAt2tz

Cat’s outta the bag! OpenAI are as strategically awful as many have speculated. “The Authors Guild has highlighted some of the now public comments made by OpenAI executives in the case. "The better we do on GPT-X, the more worried genre fiction authors will become about us substituting for them on Amazon,” OpenAI policy director Jack Clark stated in testimony given in May 2020. He further acknowledged that while “there will be a point where a bunch of artists express worry about what we’re doing here and we’ll likely ignore their concerns and release anyway.” __________ “…Gogineni was aware of authors’ complaints that the ‘datasets are stolen’ and that authors were losing work to AI-generated competition. Nevertheless as the brief goes on, he did not find those complaints ‘all that sympathetic’ and instead viewed them as ‘acceptable economic disruption.’ He wrote that the world would soon experience ‘the death of the reader’ as ‘machines creat[ed] slop for more machines.’” https://lnkd.in/eZgHMRuR

Added to last week’s revelations, new filings reveal a lawless determination to do virtually anything to win the AI arm’s race, including using books from a “sketchy Russian website” and releasing products “that will make people unemployed.” https://lnkd.in/ek2PaZ-u

OpenAI’s recent involvement in a legal battle over book piracy has sparked a lot of discussion, particularly around their concerns about "optics" on platforms like Hacker News. But let's dive deeper: what does this say about the broader implications for AI ethics and corporate responsibility? If a tech giant like OpenAI is worried about public perception over their use of copyrighted material, it signals a much larger issue. The ethical considerations need to catch up with the technological advancements. It's not just about the tech; it's about the societal impact and legal ramifications. What do you think? Should there be clearer boundaries set by tech companies around the use of copyrighted material in AI training data? 🤔 Let’s keep this discussion going and ensure we’re making responsible decisions in our field. Practical takeaway: if this shaves time off your workflow, that's skill value you can bring to your next project or interview.

OpenAI and Microsoft Documents Raise Questions About AI’s Impact on the Web Recently unsealed court documents from The New York Times’ case against OpenAI and Microsoft contain stark internal warnings about the consequences of AI data practices. The documents reportedly describe: • A potential “doom loop” in which AI systems could damage the web by reducing traffic and weakening the sources they rely on • The scraping of data for model training as the “largest theft of labor in human history” • Concerns that these practices could make a “complete mockery” of fair use Several of the most striking comments were attributed to Brent Hecht, Microsoft’s Director of Applied Science. Microsoft has sought to distance itself from some of his assertions, with a spokesperson indicating that the comments do not represent the company’s official position. The documents add to a broader debate over copyright, fair use, AI training data, and the long-term sustainability of online publishing. The case remains ongoing, and the statements reflect arguments and internal discussions presented in litigation—not a final legal ruling. ⚖️ The central question remains: How can AI development progress while preserving the people and businesses that create the information these systems depend on? #SuperIntelligenceNews #SuperIntelligenceNewsletter #ArtificialIntelligence #OpenAI #Microsoft #AIEthics #Copyright #FairUse #DigitalMedia #TechnologyNews

It is encouraging that some Authors are getting compensated for their work being used to 'train" Ai models. Maybe this means that next time, they will get paid in advance, and at a fair rate, when their work is being used. https://lnkd.in/dWc9J246

Microsoft Exec Called AI Scraping “The Largest Theft of Labor in Human History,” Unsealed Filings Reveal Newly unsealed court filings reportedly show that a Microsoft executive privately described AI companies’ use of online content as “the largest theft of labor in human history.” The filings also indicate that Microsoft and OpenAI scraped paywalled content from The New York Times, used it to build datasets, and internally warned that these practices could severely damage — or even “gut” — publishers. The documents offer a look into how the companies discussed AI training data and publisher concerns behind closed doors. The revelations come amid broader legal and industry debates over copyright, licensing, journalism, and whether AI developers should compensate publishers and creators whose work is used to train their systems. The filings present internal views and allegations; the wider legal questions remain unresolved. ⚖️ #SuperintelligenceNews #SuperintelligenceNewsletter #ArtificialIntelligence #AI #OpenAI #Microsoft #Copyright #AIEthics #Media #Publishing #TechNews #GenerativeAI

The Third Circuit’s decision in Thomson Reuters v. ROSS Intelligence does not say that AI training is never fair use. But it does show when an AI-training fair use defense can fail. First, the Court found the use only minimally transformative. Thomson Reuters used its headnotes to help legal researchers identify relevant judicial opinions; ROSS used those same headnotes to train a system designed to help users identify relevant judicial opinions. Second, copying the headnotes was not considered necessary to achieve ROSS’s objective. ROSS already had access to millions of uncopyrighted judicial opinions and could have used those opinions to build its training materials. The headnotes were used because they provided an easier route to constructing the dataset. In the Court’s formulation, “ease” was not a sufficient justification for copying protected expression. Third, the amount copied mattered. ROSS incorporated the full text of thousands of Westlaw headnotes into its training memoranda. Because each headnote was itself treated as a copyrightable work, the Court viewed this as copying entire works without a sufficiently transformative justification. Fourth, and perhaps most importantly for the broader AI debate, the Court found market harm at two levels. ROSS was building a competing legal-research platform and the Court also accepted that a rapidly developing market exists for licensing copyrighted materials as AI training data. In its view, unauthorized copying could usurp the copyright owner’s opportunity to participate in that licensing market. But the judgment should not be read as a general rejection of fair use for AI training. The Court expressly distinguished the litigation involving generative AI models. Referring to Bartz v. Anthropic and the pending OpenAI litigation, it noted that large language models capable of generating original responses may raise a materially different question of transformative use. ROSS’s system, by contrast, was non-generative: it did not create new expression and the evidence showed that it was designed as a commercial substitute for Westlaw.