Who owns the data AI learns from? Latin America is debating it now
Seventeen countries met on AI and copyright, and a Brazilian court ruled against Meta. Here's what's happening and why creators should care.

AI models learn from huge amounts of text, images and news. Who gets a say, and who gets paid? Latin America is starting to answer.
A regional meeting on AI and copyright
The World Intellectual Property Organization (WIPO) and the Dominican Republic’s copyright office planned to host delegations from 17 Latin American and Caribbean countries in Santo Domingo, November 10–12, to discuss copyright and AI. (Puente LATAM)
A court decision in Brazil
A Brazilian court blocked Meta from using news content to train its AI, according to a regional tech digest. (Puente LATAM) It’s one of several legal fights worldwide over whether AI companies can use news and creative work without permission or payment.
Countries competing for AI data centers
At the same time, governments want AI investment. Argentina is offering tax breaks and minimal AI regulation to attract data centers, while Brazil is tying data-center tax breaks to local access as big tech expands. (Puente LATAM)
What it means for you
- If you create content (writing, photos, music, video), the rules about using your work to train AI are being decided now. Follow your country’s copyright office.
- If you use AI at work, don’t paste in content you don’t have the right to use, and check your company’s rules.
- If you work in tech, AI policy and compliance are becoming real career paths, especially for people who know both the technology and the region.
These summaries come from the reports linked above. Legal decisions can be appealed and policies can change.