Meta CEO Mark Zuckerberg addresses the intricate copyright issues surrounding data scraping for AI training. In an interview with Alex Heath, he indicated that while Meta plans to pursue "certain partnerships" for valuable content, they will not compensate individual creators if they demand payment, similar to their approach with news outlets. "I think individual creators or publishers tend to overestimate the value of their content in the grand scheme," Zuckerberg stated. He emphasized that if creators object, Meta would refrain from using their content, but asserted that this would have minimal impact on overall outcomes.
Meta, like many AI companies, is caught in litigation over the legality of scraping data for AI model training. The company faced a lawsuit from authors, including Sarah Silverman, who alleged that Meta's Llama model was improperly trained on their copyrighted works. However, the case faced challenges, with a judge criticizing the authors' legal team.
Zuckerberg argues for a re-examination of fair use in the context of AI, stating, "When you put something out in the world, how much control do you retain over it?" The history of copyright involves determining the level of control authors have over their published work, and fair use is intended to encourage creative transformation without requiring permission.
While the legal debate continues, several AI companies have established paid partnerships with major media outlets. OpenAI, for instance, has formed alliances with various news organizations and companies like Shutterstock, while Meta has recently contracted Universal Music Group regarding AI-generated music.
Some artists are seeking unofficial methods to protect their work from being used for AI training. However, those who posted content on social media prior to the generative AI boom often face obstacles due to existing terms of service that allow companies to utilize their content. Meta has clarified that it trains its AI models on public posts from Facebook and Instagram.
Zuckerberg’s future AI content strategy appears to mirror Meta's previous responses to proposed legislation that would impose fees for linking to news articles. Historically, the company has blocked news outlets in countries like Australia and Canada when faced with such regulations. "We're a big company," Zuckerberg said. "We pay for content when it’s valuable. We won’t pay for content that isn’t." Given Meta's track record, it’s evident that creators should not anticipate monetary compensation from AI developments.
Ranking
