Copyright vs Neural Networks: Indonesia Prepares Strict Law for Google and AI Platforms

The model of free data parsing for LLM training is collapsing before our eyes. On July 17, 2026, a confirmed draft law from the Indonesian Parliament and Ministry of Law leaked online, fundamentally rewriting copyright rules in the generative intelligence era.

The document creates a highly dangerous precedent for Big Tech. For the first time at the state level, a requirement for mandatory content labeling (AI-assisted), disclosure of training sets, and crucially, a direct ban on imitating the unique signature style of authors is being introduced. If passed, the law will undermine the unit economics of hyperscalers in the region. Google, OpenAI, and other AI platforms will have to pay massive licensing fees to local creators and publishers for every byte of data fed to neural networks. This legal framework could become a template for emerging markets, forcing IT giants to move from pirated scraping to the legal procurement of datasets.

Source: DPR RI / Ministry of Law Indonesia / Reuters