OpenAI Shares Court Briefs Supporting Fair Use for AI Training in NYT Lawsuit

OpenAI has published a collection of legal briefs and supporting arguments from authors, academics, technology organisations, accessibility advocates and other groups defending the application of fair use to artificial intelligence training. The materials were submitted in the ongoing copyright litigation involving OpenAI, Microsoft and The New York Times in the US District Court for the Southern District of New York. OpenAI published the collection on October 30, 2026, under the title “Fair use benefits everyone.”

The legal dispute centres on whether AI companies can use copyrighted books, articles and other material to train large language models without obtaining individual licences from copyright holders. OpenAI argues that training models on copyrighted material can qualify as transformative fair use under US copyright law. Publishers and authors challenging the practice argue that their work was copied without permission and that AI-generated products can compete with the original content. The latest submissions bring a wider range of public-interest arguments into an increasingly important legal debate.

Key Takeaways

• OpenAI published court briefs and supporting arguments defending fair use for AI training on October 30, 2026.
• The materials relate to the consolidated copyright litigation involving The New York Times, OpenAI and Microsoft.
• Supporters include authors, academics, libraries, accessibility organisations and technology industry groups.
• The arguments say AI training creates new tools and capabilities rather than simply reproducing copyrighted works.
• Critics argue that using copyrighted material without permission can undermine publishers’ businesses and creators’ rights.
• The publication of supporting briefs does not mean the court has ruled in OpenAI’s favour.
• The outcome could influence how AI companies obtain, use and license training data.

What OpenAI Has Published

OpenAI’s article, titled “Fair use benefits everyone,” presents excerpts from submissions made by organisations and individuals supporting the application of fair use to AI training.

These submissions are known as amicus curiae briefs, commonly described as friend-of-the-court filings. They allow people and organisations that are not necessarily parties to a lawsuit to present legal arguments or explain how a court decision could affect broader interests.

The collection includes arguments from several groups:

• American Council of the Blind
• Association of Research Libraries
• American Library Association
• Consumer Technology Association
• Business Software Alliance
• Authors Alliance
• Public Knowledge
• Foundation for American Innovation
• Individual academics, writers and creative professionals

The submissions approach the dispute from different perspectives, including innovation, accessibility, education, research, competition and the future of creative industries.

Importantly, these are arguments supporting one side of a contested legal question. They are not judicial findings or proof that every use of copyrighted material for AI training is lawful.

Why OpenAI Believes AI Training Is Fair Use

OpenAI’s central argument is that training an AI model is fundamentally different from distributing copies of the original books, articles or other material used during training.

When a model is trained, it processes large collections of information to learn statistical patterns, relationships and language structures. The resulting model is designed to generate responses to new instructions rather than function as a conventional archive of its training documents.

OpenAI argues that this makes AI training transformative under US copyright law.

The company also contends that AI systems provide public benefits across several industries, including scientific research, education, accessibility, software development and creative work.

The legal question is whether those benefits, together with the purpose and effects of the copying, justify treating the training process as fair use under the relevant circumstances.

What the Supporting Organisations Are Arguing

One of the most notable aspects of the collection is the range of organisations supporting OpenAI’s position.

The American Council of the Blind, Association of Research Libraries and American Library Association argue that the case affects more than companies developing AI models.

Their submission points to people with disabilities, researchers, educators, students and institutions that depend on tools built using broad collections of information.

The argument is that restricting AI training could affect organisations that use AI products without developing the underlying models themselves.

Authors Alliance makes a different legal argument. It challenges the idea that copyright owners should automatically be able to claim market harm simply because new technologies or creative works compete with their products.

Public Knowledge draws comparisons with earlier technologies that benefited from fair use, including search engines, digital book search and accessible digital libraries.

Together, these submissions argue that copyright law should protect original expression without preventing the development of new tools that learn from information.

The New York Times’ Position in the Lawsuit

The New York Times and other publishers have argued that OpenAI and Microsoft used copyrighted journalism without permission to develop commercial AI products.

Their claims include allegations that copyrighted articles were copied during the development of AI models and that ChatGPT can sometimes reproduce or closely resemble protected material.

The publishers argue that this practice threatens the economic value of original journalism and undermines licensing arrangements that support news organisations.

They also raise concerns about AI-generated responses potentially substituting for visits to publishers’ websites and reducing opportunities to generate subscription and advertising revenue.

OpenAI disputes the central copyright allegations and argues that model training is protected by fair use.

The disagreement therefore involves two competing views of the relationship between copyright and technological innovation.

Issue	OpenAI and fair-use supporters	Publishers and authors
AI training	Transformative use of information	Unauthorised copying of protected works
Market impact	New tools create additional value	AI products can compete with original content
Innovation	Fair use supports technological progress	Licensing can support creators while enabling AI
Copyright protection	Protects expression, not every use of information	Should protect creators against unlicensed commercial exploitation
Public interest	Broad access to AI benefits	Sustainable journalism and creative industries

Why the US Government’s Position Matters

OpenAI’s published collection also includes arguments from the United States government.

The government’s submission challenges attempts to narrow fair-use doctrine in a way that would exclude AI training.

It argues that doing so could interfere with technological progress and research, including applications that help scientists and other professionals develop new capabilities.

The submission also points out that publishers themselves are using large language models for journalistic and other purposes.

This introduces an important distinction: the debate is not simply about whether AI should be used in publishing or research. It is about the legal conditions under which AI developers can use copyrighted material to build the systems that make those applications possible.

The government’s participation is significant, but its arguments do not determine the outcome. The court must independently assess the legal claims and evidence.

What Happens Next in the Copyright Battle?

The fair-use question is central to the consolidated litigation in the Southern District of New York.

The court must consider the circumstances of the copying, the purpose of AI training, the nature of the copyrighted works, the amount used and the effect on existing or potential markets.

These are the four factors traditionally considered under Section 107 of the US Copyright Act.

A decision could influence how AI companies train models and how publishers negotiate licensing agreements.

If the court accepts OpenAI’s arguments in the relevant circumstances, it could strengthen the legal basis for certain AI training practices. If it sides with the publishers, AI developers could face greater pressure to obtain licences or change how they source training material.

The implications could extend well beyond OpenAI, affecting other AI developers and companies building products on top of foundation models.

The Bigger Picture

The dispute reflects a fundamental challenge for copyright law in the age of generative AI: how to balance the rights of creators with the development of technologies that learn from existing information.

Fair use has historically helped enable technologies such as search engines, digital libraries and other information tools. AI developers argue that model training belongs within this tradition.

However, creators and publishers face legitimate questions about compensation, control and competition. A system that allows AI companies to use large collections of copyrighted material without meaningful safeguards could change the economics of journalism, publishing and creative work.

The eventual legal outcome will help determine how much freedom AI developers have to use existing works and where copyright owners retain control.

Looking Ahead

The next major development will be the court’s treatment of the competing legal arguments and evidence. OpenAI’s publication of supporting briefs increases public visibility into the case, but it does not settle the fair-use question. The court will need to evaluate the specific uses of copyrighted material rather than treating every AI training process as automatically lawful or unlawful.

For the technology industry, the case could shape future AI development costs, licensing agreements and access to training data. For publishers and creators, it could influence how their work is protected and monetised as generative AI becomes more widely used. The decision could become an important reference point for the relationship between copyright law and artificial intelligence.

Frequently Asked Questions

1. What did OpenAI publish on October 30, 2026?

OpenAI published excerpts and links to court briefs supporting the application of fair use to AI training. The submissions came from authors, academics, libraries, accessibility advocates, technology organisations and other supporters.

2. What is fair use in US copyright law?

Fair use is a legal doctrine under Section 107 of the US Copyright Act. It permits certain uses of copyrighted material without permission, depending on factors such as purpose, nature of the work, amount used and market impact.

3. Why is The New York Times suing OpenAI?

The New York Times alleges that OpenAI and Microsoft used its copyrighted journalism without permission in developing AI systems. The publishers also raise concerns about reproduction of their work and competition with their original content.

4. What is an amicus curiae brief?

An amicus curiae brief is a submission from an individual or organisation that is not necessarily a party to a lawsuit but wants to provide information or legal arguments that may help the court.

5. Does the publication of these briefs mean OpenAI has won the case?

No. The briefs present arguments supporting OpenAI’s position. The court has not established through these submissions that AI training is automatically fair use.

6. Could the case affect other AI companies?

Yes. A decision on AI training and fair use could influence other companies developing foundation models, including their data acquisition strategies, licensing arrangements and legal risks.

7. Could publishers receive compensation for AI training?

That depends on licensing agreements, legal decisions and the specific circumstances of each use. The case could influence whether certain training practices require permission or payment.