NYT accuses Microsoft of building copyright‑infringing supercomputer for OpenAI

NYT slams Microsoft for building a copyright‑infringing supercomputer

The New York Times published on Tuesday a report accusing Microsoft of constructing a supercomputer dedicated to training OpenAI models that allegedly ingested massive amounts of copyrighted material without proper licenses. According to the article, the infrastructure—based on Nvidia H100 GPU clusters—was fed billions of pages of books, articles, and digital artwork, many still under legal protection.

Technical impact of the allegation

From a technical standpoint, such a supercomputer enables OpenAI to slash model‑training time from months to weeks. However, indiscriminate data ingestion can introduce bias and infringe intellectual property rights, exposing Microsoft to lawsuits and forcing the implementation of stricter compliance filters. The report notes that the hardware architecture lacks provenance‑tracking mechanisms, making post‑hoc audits difficult.

Market repercussions

For investors and competitors, the accusation raises questions about the sustainability of a business model that relies on unlicensed data at scale. Media companies, publishers, and creators may push for licensing agreements or regulatory limits on the use of protected material in AI training. Simultaneously, the controversy could accelerate adoption of “clean‑room” AI solutions that guarantee legally sourced data, creating new market opportunities.

10Dobro Prod’s stance

At 10Dobro Prod, we believe responsible innovation is essential for the credibility of the AI industry. While we monitor the legal developments closely, we reaffirm our commitment to using only licensed sources in audiovisual content generation projects. Transparency in training processes and strict adherence to copyright law are competitive and ethical differentiators for our clients and partners.