The US Copyright Office has thoughts on how AI is trained. Big Tech may not like it.

The US Copyright Office released a report questioning whether Big Tech companies' use of copyrighted content to train AI models falls under fair use, potentially challenging practices by companies like OpenAI. The report suggests that while AI models used for research may be transformative, commercial use of copyrighted works without permission may not qualify as fair use. This stance could lead to legal challenges for AI companies heavily reliant on vast datasets. The report distinguishes between AI models used for research and those for commercial purposes, emphasizing that outputs competing with existing markets are less likely to be considered transformative.
A day after the report's release, President Donald Trump fired US Copyright Office Director Shira Perlmutter, sparking controversy and allegations of political interference. Critics, including New York Rep. Joe Morelle, argue that Perlmutter's dismissal was an unprecedented power grab linked to her refusal to support Elon Musk's AI initiatives. Trump's close ties with tech billionaires like Musk, Zuckerberg, and Altman, who have supported his administration's efforts to curb federal spending, further intensify the debate over the role of Big Tech in shaping AI policy and the potential impact on copyright law.
RATING
The article provides a timely and relevant exploration of the intersection between AI technology and copyright law, highlighting the potential legal challenges and implications for various stakeholders. It effectively uses the US Copyright Office's report as a basis for discussing fair use and AI training. However, the story could benefit from more balanced perspectives and additional primary sources, particularly regarding the political elements and the firing of Shira Perlmutter. While the article is generally clear and engaging, the inclusion of speculative political motivations without sufficient evidence may affect its credibility. Overall, the story is informative but would be strengthened by more comprehensive sourcing and transparency.
RATING DETAILS
The story accurately reflects the general concerns about AI training and copyright issues, citing the US Copyright Office's report on AI and fair use. The mention of Big Tech companies using copyrighted material for AI model training is a known practice, though specifics could be further verified. The report's discussion on the transformative nature of AI training and its potential copyright implications is well-captured, aligning with the US Copyright Office's stance. However, the claim about Shira Perlmutter's firing by Donald Trump and the suggested political motivations require more substantiated evidence, as the timing and context could be speculative without direct sources or official statements.
The article presents perspectives from both the US Copyright Office and AI companies, outlining the tension between creators' rights and AI development needs. However, it leans towards highlighting the creators' objections and the potential legal challenges faced by AI companies. The narrative about Trump's involvement and the firing of Shira Perlmutter introduces a political angle, which could skew the reader's perception without additional viewpoints from the involved parties, such as the White House or AI companies directly.
The article is generally clear and structured, presenting the issues surrounding AI training and copyright in a straightforward manner. However, the inclusion of political elements, such as Trump's involvement, might confuse readers if not well-explained or contextualized. The language is accessible, but the narrative could benefit from clearer delineation between factual reporting and speculative commentary.
The article references the US Copyright Office's report, a credible source for discussions on copyright law. However, the story lacks direct quotes or statements from involved parties like AI executives, legal experts, or government officials. The absence of such primary sources diminishes the reliability of some claims, particularly those related to political motivations and the firing of Shira Perlmutter.
The article does not provide detailed information about how the claims were sourced or verified. While it references the US Copyright Office's report, it does not clarify the methodology or context behind other claims, such as the firing of Shira Perlmutter. The lack of explicit sources or evidence for some of the more controversial claims affects the transparency of the reporting.
Sources
- https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-3-Generative-AI-Training-Report-Pre-Publication-Version.pdf
- https://www.copyright.gov/ai/
- https://chatgptiseatingtheworld.com/2025/05/10/us-copyright-issues-pre-publication-version-of-3rd-report-on-ai-training-and-fair-use-ai-training-is-transformative-but-degree-depends-on-how-ai-functions-supports-new-market-dilution-theory-of-ha/
- https://copyrightlately.com/copyright-office-ai-report/
- https://ppc.land/us-copyright-office-releases-major-ai-training-report-amid-intensifying-copyright-debate/
YOU MAY BE INTERESTED IN

Public comments to White House on AI policy touch on copyright, tariffs
Score 6.2
OpenAI’s models ‘memorized’ copyrighted content, new study suggests
Score 7.2
FTC Commissioner Melissa Holyoak working remotely from Utah — despite Trump’s RTO order and DOGE clampdown: sources
Score 5.4
At CPAC, the MAGA base is skeptical of Trump’s Big Tech alliance | CNN Politics
Score 6.2