Microsoft memos reveal AI's 'destructive loop' of stealing content, threatening creators and jobs. A landmark trial may redefine copyright laws for generative AI.
Declassified Microsoft memos reveal AI strategy
Internal memos from Microsoft, declassified in a federal trial in New York on September 17, 2026, describe the company's artificial intelligence strategy as "a theft of unprecedented scale." These documents, submitted by The New York Times to a case against Microsoft and OpenAI, reveal concerns about a self-destructive cycle where chatbots absorb human-produced content while depriving creators of income.Microsoft official warns of destructive loop
A memo written by Brent Hecht, director of applied sciences at Microsoft, refers to this as a "destructive loop," noting a "very unusual" situation where a commercial product threatens the viability of those who provide it with data. The "content supply chain" remains vital for training AI, even though its deployment makes it inoperable, according to the memo.Public perception of AI as large-scale theft
Another memo suggests that millions of people may perceive the capture of their work as "a theft of unprecedented scale," as reported by Futurism. This trial, which involves eleven publishers, could redefine the legal boundaries of the entire industry. Artificial intelligence tools summarize news articles without directing readers to the original sites, depriving newsrooms of clicks and advertising revenue.Creative job losses and industry impact
The American Bureau of Labor Statistics records the loss of over 200,000 creative jobs in four years, with approximately 50,000 of these job losses occurring in the last year. Half of the eliminated positions are concentrated in cinema and audiovisual sectors. Musical streaming platforms see algorithm-generated tracks flooding in at the expense of artists.Education and labor affected by AI
These vicious circles also affect the educational world. Students delegate their homework to chatbots while their teachers design exercises with the same software. The Brookings Institution concluded in January 2026 that the risks of generative AI in the classroom outweigh the benefits. Its studies, conducted in fifty countries, also point to a decline in reading and critical thinking skills.Judicial review and future implications
Judge Sidney Stein oversees the case in New York and examines the requests for summary judgment. His decision could create a global precedent on copyright in the face of generative AI. However, without a framework imposed by the judiciary, the loop will tighten. Creators will stop producing the fresh material these systems need. The industry will lose its raw material.