> ## Content Index
> Fetch the complete content index at: https://nordisk-ai.ghost.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# OpenAI Gives Authors Access to Training Data in Copyright Case
- URL: https://nordisk-ai.ghost.io/openai-gives-authors-access-to-training-data-in-copyright-case/
- Published: 2024-11-02T07:28:46.000Z
- Updated: 2026-08-27T17:21:05.000Z
- Description: Authors suing OpenAI are now getting the chance to review the AI company's training data, in a closely watched copyright case where they claim their work was used without permission or compensation.
- Author: Anders Hagen
- Tags: Legal AI, AI, OpenAI, Sweden, Sweden, copyright law, #Migrated-1787841634512, #wp, #wp-post, #Import 2026-08-27 14:41, #Import 2026-09-07 13:41

Authors suing OpenAI are now getting the chance to review the data used to train the AI model, in an effort to determine whether their copyrighted work has been used without permission or compensation.

For the first time, OpenAI is providing insight into its training data as part of an ongoing legal dispute in which several authors accuse the company of using their work without consent. In a joint agreement announced on Tuesday, guidelines were set out for how the authors' representatives will be allowed to review data that may contain their work.

### Background to the Lawsuit

The agreement comes after several authors, including Sarah Silverman, Paul Tremblay, and Ta-Nehisi Coates, filed lawsuits against OpenAI. They allege the company used unauthorized amounts of their books to train ChatGPT, resulting in unlawful use of their material in AI-generated responses. Earlier in July, a court dismissed some of the claims against OpenAI, but the authors' central claim of direct copyright infringement remains.

### Fair Use Arguments

Many AI companies have previously denied copying works outright, instead claiming their training models build parameters based on such works to define how text and images should be constructed. OpenAI may later defend itself by arguing that its use of published works falls under "fair use," a legal principle that permits the use of copyrighted material if it's transformative.

### Authors' Claims of Unauthorized Data Collection

OpenAI has previously acknowledged that its AI was trained on large, publicly available datasets that may include copyrighted material. But since last year, the company has stopped disclosing details about its training data to protect itself from legal risk and competition. The authors argue that ChatGPT has generated detailed analyses of their work, suggesting their books were used in training. They also claim OpenAI downloaded hundreds of thousands of books from illegal sources.

### Strict Security Measures During Review

Under the agreement, the authors' representatives will get access to the training data at OpenAI's office in San Francisco. The review will take place on a secure computer with no internet connection, and everyone taking part must sign nondisclosure agreements and show identification. No recording devices are allowed in the room, and notes may only be taken on a computer under close supervision.

The review is tightly restricted. OpenAI only permits handwritten notes or electronic notes on the provided computer, and no copies of the training data may be made.

### Similar Lawsuits Against Meta

The authors' legal team from the Joseph Saveri Law Firm is also leading similar copyright cases against Meta. There too, the court has expressed criticism of the lawyers' handling of the case, particularly over the lack of depositions and the short time remaining for fact-finding.