“ #Facebook and #Instagram owner Meta Platforms has indicated it won’t be signing on to the European Union’s voluntary 
#AI Code of Practice, which includes restrictions on how AI companies can collect #Copyrighted content.”
“Suno, for those of you not familiar, is an #AI #SongGenerator: enter a text prompt (such as “a jazz, reggae, EDM pop song about my imagination”) and a song comes back. Like many #GenerativeAI companies, it is also being sued by all and sundry for ingesting #copyrighted #material. The parties in the suit — including major labels and the #RIAA — don’t have a smoking gun, since they can’t directly peek at Suno’s #TrainingData. But they have managed to generate some suspiciously similar-sounding AI generated materials, #mimicking (among others) “Johnny B. Goode,” “Great Balls of Fire,” and Jason Derulo’s habit of singing his own name.
#Suno essentially admits these songs were #regurgitated from #copyrighted source material, but it says such use was legal. “It is no secret that the tens of millions of #recordings that Suno’s model was trained on presumably included recordings whose rights are owned by the Plaintiffs in this case,” it says in its own legal filing. Whether AI training data constitutes fair use is a common but unsettled legal argument, and the plaintiffs contend Suno still amounts to “pervasive #illegal #copying” of artists’ works.”
#NYA / #music / #ElizabethLopatto / #amazon / #DataTheft <https://neilyoungarchives.com/news/3/article?id=Music%20-%20Amazon%20is%20blundering%20into%20an%20AI%20copyright%20nightmare>
@Techmeme
if the AL #bigCorp can #watermark images created from free account then they should have no problem just watermarking all images created with #copyrighted material.
Will any court get on this?
#OpenAI declares #AI race “over” if #training on #copyrighted works isn’t fair use
OpenAI is hoping that Donald Trump's AI Action Plan, due out this July, will settle #copyright debates by declaring #AItraining fair use—paving the way for AI companies' unfettered access to training data that OpenAI claims is critical to defeat #China in the AI race.
#fairuse #Trump
#Zuckerberg gave #Meta's #Llama team the OK to train on copyrighted works, filing claims, using a dataset of pirated e-books and articles.
Zuckerberg approved Meta’s use of a dataset called #LibGen for Llama training.
LibGen provides access to #copyrighted works from publishers including Cengage Learning, Macmillan Learning, McGraw Hill, and Pearson Education. LibGen has been sued, ordered to shut down, and fined tens of millions of dollars for #copyrightinfringement.
https://techcrunch.com/2025/01/09/mark-zuckerberg-gave-metas-llama-team-the-ok-to-train-on-copyrighted-works-filing-claims/
It’s January, which means another batch of #copyrighted work is now public domain
#publicdomain #copyright
→ Former OpenAI Researcher Says the Company Broke Copyright Law
https://www.nytimes.com/2024/10/23/technology/openai-copyright-law.html
“Suchir Balaji spent nearly four years as an artificial intelligence researcher at OpenAI. [A]fter the release of ChatGPT in late 2022, he thought harder about what the company was doing. He came to the conclusion that OpenAI’s use of copyrighted data violated the law and that technologies like ChatGPT were damaging the internet.”
NVIDIA Reportedly Scraped Copyrighted, Academic, And Non-Commercial Content For AI Training #academic #ai #airobotics #content #copyrighted #noncommercial #nvidia #training
https://www.lowyat.net/2024/328414/nvidia-scraped-content-ai-training/
@marcan @bunnie precisely that!
That's also why AI can't violate nor create copyright becaudr only natural persons can create any intellectual property that is protectable and if we'd claim that #AI could commit "Copyright Infringement" we'd also allow the #Copyrightmafia to hold everyone as perpetual #DebtPeon if tuey ever used #copyrighted materials to learn anything (i.e. no artist learning from a copyrighted score would be able to make permissively licensed music!)
#Apple et. al. use all the other mixes of IP to go after clones, besides absurd patents (i.e. #MagSafe) they mostly use exact proportions and dimensional accuracy to combat lookalike devices.
As for the #ISA #patents issue: #RISCv was specifically designed to workaround this issue in #academia and not have professors constantly violate #NDA|s on top of thise...
@grmpyprogrammer that's because it's legally and technically not the same, otherwise half of the #Millenials in #Germany would be perpetual #DebtPeons to #JKR for their #English skills...
https://felixreda.eu/2021/07/github-copilot-is-not-infringing-your-copyright/
Am I an #accelerationist when I hope that #LLM|s being trained on #copyrighted content will lead to a new copyright system that is fairer for creators and consumers?
I’m pretty sure I am naive.
A good set of points from @gruber. I totally agree that the public web is a reasonable place for #AI #LLM models to “learn” from. “Learning” whether from a human or a computer has always been within the realm of proper use of #copyrighted material. The one key thing is that AI needs to be safeguarded so it doesn’t rip off the material it learns from. But, humans are also apt to plagiarize, as I and anyone else who has ever taught students in a classroom can surely attest to, so this is not a new issue. https://daringfireball.net/2024/06/training_large_language_models_on_the_public_web
Needless to say, if "#SourceAvailable = #OpenSource" were true, than any leaked #sourcecode would mean it can't be #copyrighted...
Does anyone have thoughts on #Archivedotorg / #WaybackMachine archiving your #WordPress etc #domain containing original content that you authored? Is this a #copyright #law violation in the same vein as Archive[dot]org posting a #copyrighted song or movie?
Thx! I'm interested in light of discussions about content being sold to #AI companies sans author permission. Also curious: has anyone had success getting Wayback to remove or exclude #writing from its site with or without a copyright claim?
“It looks like your #blocker is attempting to interfere with the intended operation of this site. Support our writers and our #copyrighted #content by allowing our site to function as we intended. Please disable your #blocker and add us to your #allowlist.”
#OpenAI admits it's impossible to train #generativeAI without #copyrighted materials
https://www.engadget.com/openai-admits-its-impossible-to-train-generative-ai-without-copyrighted-materials-103311496.html?src=rss
#AI companies would be required to disclose #copyrighted training data under new bill - The Verge
The AI Foundation Model Transparency Act aims to make it clear if AI models used #copyright data for training.
Rather than remove #copyrighted material from #ChatGPT’s training #dataset #chatbot’s creator #OpenAI CEO #SamAltman offering 2 cover its clients’ legal costs 4 copyright infringement suits. We can defend our customers pay costs incurred if u face legal claims around #copyrightinfringement applies to #ChatGPT Enterprise & API. Compensation offer #CopyrightShieldapplies to users of business tier, ChatGPT Enterprise, dev using ChatGPT’s application programming interface. https://www.theguardian.com/technology/2023/nov/06/openai-chatgpt-customers-copyright-lawsuits
#AI companies have all kinds of arguments against paying for #copyrighted content - The Verge
♬ 
