According to a court document unblocked on October 16, Microsoft had ordered employees to “keep secret” its website scraping project; the project collected artificial intelligence training data for OpenAI. In this project, codenamed “Project Mango,” OpenAI paid Microsoft to develop and operate a web crawler to copy web content; this content was then used to train models including GPT-4o. The news publisher said in a summary judgment motion that was partially unblocked.

According to a court document unblocked on October 16, Microsoft had ordered employees to “keep secret” its website scraping project; the project collected artificial intelligence training data for OpenAI. In this project, codenamed “Project Mango,” OpenAI paid Microsoft to develop and operate a web crawler to copy web content; this content was then used to train models including GPT-4o. The news publisher said in a summary judgment motion that was partially unblocked.