The New York Times has filed a copyright lawsuit against Microsoft and OpenAI, the owners of ChatGPT, making it the first major US media organization to do so. It alleges that significant amounts of copyrighted work have been infringed upon to “develop their models and tools.”
OpenAI is a type of large language model (LLM), and its performance is refined by analyzing substantial amounts of data typically sourced from ChatGPT users online.
The lawsuit provides explicit examples of how its generative AI (GenAI) tools produce output that “recites Times content verbatim,” carefully analyses it, and “mimics its expressive style.” Artificial intelligence models that can feed back articles, particularly those behind a paywall, pose a financial risk to newspapers. This suit claims that such tools “undermine” the relationship with current subscribers and “deprive” the company of potential subscribers, license payments, and advertising revenue.
The New York Times is seeking unspecified billion-dollar damages from both companies for allegedly profiting from the “exploitation and misappropriation” of its intellectual property.
Earlier this year, it raised its concerns with OpenAI and Microsoft in an attempt to reach a commercial agreement that would be mutually beneficial. However, after months of negotiations, there’s been no resolution.
This recent development highlights how media companies have had a tense relationship with AI, fraught with litigation.
In the past year, OpenAI has faced a slew of copyright infringement lawsuits. Among other bestselling authors, George RR Martin sued the company, citing “systematic theft on a mass scale.” This was followed by other suits from writers and journalists, namely Julian Sancton and New Yorker writer Jia Tolentino, claiming similar reasons.
Comedian/writer Sarah Silverman also brought legal action against OpenAI. In the same month, an open letter signed by over 15,000 writers, including Margaret Atwood, called for AI companies to “obtain consent, credit, and fairly compensate writers” for the use of their work. OpenAI has yet to resolve any of these matters.
This rising tension isn’t just with publishing. Stock photo provider Getty Images filed a lawsuit against Stability AI Inc. for misusing over 12 million of its photos to train the company’s image generation tool. Again, the system was fed licensed content without permission to produce more accurate imagery from user prompts.
The point of contention is how GenAI tools - which instantly produce human-like text, imagery, and code - use data for free without permission. Training models this way could render content like news articles useless, lowering the media’s negotiating power and posing a severe threat to an already struggling industry.
However, AI supporters argue that the availability of open-source information falls within copyright laws and is, therefore, “fair use.” But by answering questions directly, tools like ChatGPT become an obvious replacement for original material. When facing trial, AI companies will need a strong rebuttal in their “fair use” argument.
Some outlets have welcomed talks with global tech companies to finalize an agreement over how news content will be used to refine AI technology.
According to industry insiders, Microsoft, Google, Adobe, and OpenAI have been in conversation with media executives regarding copyright concerns over AI products like image generators and chatbots. Sources say that talks with at least one of the tech companies have been with outlets including The Guardian, News Corp, The New York Times, and Axel Springer.
This could result in deals that require a separate subscription-style fee when using content to develop AI technology. However, these are still in the early stages.
Publishers also put forward pricing models that consist of an annual fixed payment for unlimited use of news content in AI learning tools. Previously, Mathias Döpfner, CEO of Axel Springer, saw this as a secondary option because it would be difficult for smaller or local outlets to benefit from it. But Axel Springer has since signed a landmark agreement with OpenAI to use content from its outlets, including Politico and Business Insider.
It’s difficult to ignore the revolutionary nature of this technology. While the scope of opportunity is virtually endless, tools like ChapGPT will likely upend multiple sectors in the coming years. Other industries will soon be forced to pay attention.
Editor: Kaiyah Ellison
Share This Post On
Leave a comment
You need to login to leave a comment. Log-in