The Intercept, along with several other media outlets, has filed a lawsuit against OpenAI and Microsoft for using journalists’ work without permission or credit to train their AI language model, ChatGPT. This case highlights the vulnerability of digital news outlets to AI grifters. While larger publications and Hollywood stars have also raised concerns about OpenAI’s practices, digital outlets face unique challenges in protecting their work.
The Intercept’s lawsuit, filed in February, alleges that OpenAI and Microsoft violated the Digital Millennium Copyright Act (DMCA) by using copyrighted stories without paying licensing fees or acknowledging the authors. The Intercept’s general counsel, David Bralow, argues that OpenAI and Microsoft are taking advantage of the hard work and resources invested by online news outlets, aiming to “free ride” on their efforts.
OpenAI and Microsoft have attempted to dismiss The Intercept’s lawsuit, just as they denied casting an actor with a voice similar to Scarlett Johansson’s for their virtual assistant. Last month, they filed motions to dismiss the case, which will be argued in federal court in June.
OpenAI has entered into licensing agreements with various press outlets in the past year, including the Associated Press, Le Monde, the Financial Times, and Axel Springer. However, many other outlets have sued OpenAI for copyright infringement. The New York Times, the Chicago Tribune, the New York Daily News, and six other daily papers owned by Alden Global Capital have all filed lawsuits. Digital outlets Raw Story and AlterNet have also filed a separate lawsuit.
All plaintiffs, both traditional and digital outlets, have noted that their websites appear in OpenAI’s lists of scraped pages used to train ChatGPT. The Intercept’s website is included in one of the datasets containing text from over 6,400 pages from their domain.
OpenAI and Microsoft argue that The Intercept failed to identify specific articles that were used to train ChatGPT. However, The Intercept counters that only OpenAI and Microsoft have access to that information unless the court allows the case to proceed into discovery.
Unlike print publications, digital outlets face challenges in copyright protection due to the lack of a bulk registration process. Print publications can easily register their works with the U.S. Copyright Office, but online-only outlets must register each article individually, which is costly and time-consuming. The Copyright Office is considering a new registration process for news websites, but it is still under review.
Unable to rely on traditional copyright infringement claims, The Intercept has turned to novel arguments under the DMCA. The DMCA prohibits the intentional removal of copyright management information and the distribution of works without this information. The Intercept alleges that OpenAI and Microsoft violated both provisions by training ChatGPT on journalists’ articles without proper attribution.
Next month, the district court will decide whether The Intercept’s lawsuit will proceed. If dismissed, OpenAI can continue using digital outlets’ work without compensation. This case raises important questions about the ethical use of AI in the media industry and the need for stronger protections for digital outlets’ intellectual property.
