AI NEWS: Judge rules using Copyrighted Material to Train AI is FAIR USE

@mentorsoft_jay

8.9K followers · usually 1.4K views a post

Views

7.5K

Likes

232

Comments

73

Shares

18

Selling · SaaS › AI Tool

Mentorsoftseen in 14 videos →https://mentorsoftai.com ↗

Why it worked

The video addresses a highly relevant and contentious topic in the AI and creative industries, leveraging a recent court ruling to generate discussion and engagement. The creator presents complex legal information in an accessible, conversational format, encouraging viewer interaction through a direct call to action.

Summary

A recent court ruling determined that while pirating books to train AI is illegal, using copyrighted books obtained legally for AI training is considered fair use. This decision impacts AI companies like Anthropic, potentially allowing them to train models without explicit permission or compensation, setting a new precedent in AI development.

Structure

1AI News: Judge rules on copyright and AI training.
2Lawsuit against Anthropic regarding book data.
3Court ruling: Pirating books is illegal, but fair use for training if obtained legally.
4Implications for AI companies and authors.
5Future outlook and call to action.

Original caption

6/24

More from @mentorsoft_jay

Their other posts in the index, biggest breakout first.

Product placement

The video discusses the company Anthropic and its AI model Claude, and mentions the datasets Books 3 and LibGen. The creator's bio link points to Mentorsoft, an AI solutions company, suggesting a potential indirect promotion or a broader context of AI industry news.

Call to action

Let me know what you guys think about this in the comments below. And don't forget to follow for your daily AI news.

On-screen text

AI NEWS: Judge rules using Copyrighted Material to Train AI is FAIR USE

Transcript

A California judge just made a landmark ruling on AI and copyright. And it's a double-edged sword for authors. In a lawsuit against Anthropic, the creators of Claude AI, the court ruled that pirating books to train AI is likely illegal, but training AI on copyrighted books, even without permission, is considered fair use if the data was obtained legally. The case stems from Anthropic scraping, pirating data sets like Books 3 and LibGen to build its internal "Forever Library" of over 7 million books. Judge also blasted the company for stealing copyrighted works instead of buying them, stating that the act of creating this internal pirate library violates copyright law and will go to trial. However, he also said that the actual use of the data for training Claude was exceedingly transformative and likely protected under fair use. That means as long as AI companies get books legally, they can train on them without permission or compensation. Well, Anthropic won't escape liability for pirating, it dodges accountability for using the data to train its models. More lawsuits are coming, but for now, the message is clear. Stealing books is illegal, but using their content to teach machines that may be just the new normal. Let me know what you guys think about this in the comments below. And don't forget to follow for your daily AI news.