Gracenote sues OpenAI for copyright infringement over metadata used in AI training

Gracenote Sues OpenAI Over Unauthorized Metadata Use

Nielsen's Gracenote filed a lawsuit in the Southern District of New York on Tuesday alleging copyright infringement against OpenAI for the unauthorized use of data in AI training. The metadata company claims OpenAI scraped and used its proprietary entertainment metadata to train large language models that power ChatGPT without permission or compensation 1

1

2

2

. The complaint, filed by Susman Godfrey lawyers, marks a significant development in AI copyright cases as it targets not just the content used for training but also the relational framework and structure of a dataset that makes Gracenote's information valuable to enterprise clients 2

2

.

Novel Legal Ground in Intellectual Property Protection

What sets this case apart from other AI copyright cases is Gracenote's focus on the proprietary sequence and framework it uses to connect metadata. Gracenote employs more than 1,000 editors who painstakingly source, research, write, and curate content descriptions and identifiers for movies, TV programs, and other media 3

3

. The company's entire Programs Database, including its metadata and the relational map editors use to connect that data, is registered with the U.S. Copyright Office 2

2

. This legal protection allows Gracenote to seek both statutory damages and actual damages for the alleged infringement.

Evidence of Copied Content Descriptions

The lawsuit provides concrete examples where ChatGPT reproduced Gracenote's work with minimal prompting. When asked to describe HBO's "Game of Thrones," ChatGPT generated a near-exact copy of Gracenote's editor-created descriptor: "the depiction of two power families -- kings and queens, knights and renegades, liars and honest men -- playing a deadly game of control of the Seven Kingdoms of Westeros, and to sit atop the Iron Throne" 2

2

. Gracenote was able to prompt ChatGPT to create copies of its identifiers and content descriptions for popular shows including "Breaking Bad," "The Office," and "Saturday Night Live" 3

3

.

Failed Licensing Attempts and Business Threats

Gracenote CEO Jared Grusd stated that OpenAI "chose to use decades of our proprietary work without permission to build and sell its models" 3

3

. The complaint reveals that Gracenote reached out to discuss a licensing agreement with OpenAI "many times over an extended time period" but the AI company "rebuffed or ignored every single attempt to do so" 2

2

. This stands in contrast to Gracenote's successful licensing deals with other AI ventures from Samsung and Google 1

1

. The unauthorized use of data threatens Gracenote's core business with media content distributors such as smart TV providers, who could potentially use the metadata scraped by OpenAI to build competing products without compensating Gracenote 2

2

.

OpenAI's Response and Industry Implications

An OpenAI spokesperson defended the company's position, stating that its AI models "empower innovation, and are trained on publicly available data and grounded in fair use" 3

3

. Gracenote is requesting an unspecified amount of monetary damages and an injunction blocking OpenAI from using its data 3

3

. Grusd emphasized that "being pro-AI and anti-theft aren't contradictory; they are the only sustainable path forward" 2

2

. This case could set a new precedent for how data providers across industries protect their intellectual property, particularly regarding the framework and structure behind datasets. With many pending cases between media companies and AI firms, the outcome of this lawsuit may influence how courts evaluate the boundaries between innovation and infringement in the licensing markets for AI training data.