Microsoft Called AI Training on News "Biggest Theft in History 2026

In a startling revelation from the ongoing legal battle between the New York Times and OpenAI, internal Microsoft documents have surfaced in which a company director described the use of news articles to train AI models as "the biggest theft in history." The phrase, written in 2023 by Brent Hecht, Microsoft's director of applied science, underscores that key figures within both Microsoft and OpenAI were fully aware of the controversial nature of their data practices.

Background: The New York Times Lawsuit and Its Implications

In late 2023, the New York Times sued OpenAI and Microsoft, alleging that the companies used its articles to train AI models without authorization or compensation. Eleven other media outlets later joined the lawsuit. Now, as part of the discovery process, internal documents have been made public, revealing candid discussions within both companies about the ethical and legal implications of their actions.

The documents paint a picture of an industry grappling with the consequences of building powerful AI systems on the backs of copyrighted content. They also raise questions about how tech companies justify their use of such data and what it means for the future of journalism.

What Microsoft's Internal Documents Reveal

The documents show that Brent Hecht, in a 2023 internal note, warned that "millions of people around the world will soon consider the fact that large models 'absorb' all their work to be an astonishing theft of unprecedented proportions." He also expressed concern about the impact on the journalism industry, stating that AI models "are a product that destroys their supply chain."

These statements indicate that Microsoft was not oblivious to the potential backlash. Instead, they suggest a calculated awareness that the company was pushing boundaries—and that the consequences could be severe.

OpenAI's Awareness: Early Warnings Ignored?

OpenAI was also aware of the risks. In 2020, long before ChatGPT's launch, Jack Clark (then policy director at OpenAI) noted in internal communications that training models on copyrighted material could lead to legal challenges. The newly released documents confirm that both companies understood the stakes but proceeded anyway.

This raises ethical questions about the balance between innovation and respect for intellectual property. While AI companies argue that training on publicly available data is fair use, critics contend that it amounts to exploitation.

Key Takeaways from the Revelations

• Internal awareness: Microsoft and OpenAI executives explicitly acknowledged the controversial nature of using news articles for AI training.
• Legal jeopardy: The documents strengthen the New York Times' case by showing intent and knowledge of potential wrongdoing.
• Industry-wide practice: The revelations suggest that other AI companies may have similar internal discussions.
• Impact on journalism: The documents highlight concerns about AI's threat to the news industry's economic model.
• Public perception: The phrase "biggest theft in history" could sway public opinion against AI companies.

Comparing the Stakes: Microsoft vs. OpenAI

While both companies are defendants in the lawsuit, their roles and exposure differ. Microsoft, as a major investor and partner, has integrated OpenAI's models into its products, potentially increasing its liability. OpenAI, as the creator of the models, faces direct allegations of copyright infringement. The table below summarizes their positions. The background to this story is set out in Ai Video: latest articles. The same theme surfaces in OpenAI Co-Creator's New AI Model Jev: 400x Cheaper, No Text, which takes a different angle.

What This Means for the Future of AI and Copyright

The revelations could have far-reaching implications. If the court finds that Microsoft and OpenAI willfully infringed copyright, damages could be substantial. Moreover, the case may set a precedent for how AI companies can use copyrighted material in the future. A related thread runs through AI Is Minting Millionaires Fast in Silicon Valley — But Many Don't Know What. Anyone following the wider picture can start with China Leads AI Adoption: 80% Use It Weekly vs 54% in US.

For journalists and content creators, the documents validate their concerns about AI's impact on their livelihoods. They also underscore the need for updated copyright laws that address the unique challenges posed by generative AI. Earlier coverage in Wind Turbine Blade Recycling Breakthrough: Blades Turned Into Asphalt shows how the argument developed. A closer look at the same subject appears in Microsoft Called AI Training on News "Biggest Theft in History.