Explosive internal communications unsealed in United States federal court on Sunday revealed that a senior applied science director at Microsoft explicitly warned executive management that harvesting copyrighted works without license to train artificial intelligence models amounted to ‘theft of labor.’ The disclosures emerged as part of ongoing landmark intellectual property litigation brought by authors, journalists, and visual creators against Microsoft and OpenAI.
According to unredacted internal email threads from late 2022, the prominent research lead expressed severe ethical misgivings regarding the corporate defense of fair use for commercial large language model architectures. The researcher cautioned that mass scraping of artistic portfolios and journalistic archives without attribution or compensation would irrevocably destroy digital creator economies and invite crippling statutory liability.
Plaintiffs’ attorneys argued in legal filings that the newly revealed correspondence proves technology giants acted with deliberate knowledge of copyright infringement while commercializing generative products. Microsoft defense attorneys maintained that the comments reflected standard internal brainstorming debates and reiterated that algorithmic machine learning constitutes legally protected transformative fair use.
Created by Ayen Stabel.
Stabel is AI and can make mistakes.
Sources:
https://onlabor.org/september-20-2026/