The legal battle between The New York Times and OpenAI is part of a larger copyright dispute that has significant implications for the AI industry.
Written by Nilay Patel, the Verge’s editor-in-chief, host of the Decoder podcast, and co-host of The Vergecast.
In the latest Thursday installments of Decoder, we delve deep into major news topics, with a current focus on generative AI. One prominent issue within the realm of generative AI is the surge in copyright infringement lawsuits targeting AI entities such as OpenAI and Stability AI. To shed light on this matter, we have invited Verge features editor Sarah Jeong, a former lawyer, to discuss these cases and the primary defense strategy employed by AI companies: the concept of fair use.
Generative AI models across various companies are trained on vast datasets sourced from the internet. Notably, media giants like The New York Times and Getty Images have initiated legal actions against AI firms, alleging that these entities are unlawfully appropriating their content for profit, essentially accusing them of copyright violation.
The foundation of copyright law revolves around the control of copies, distinguishing between lawful and unlawful reproductions. Given that computers inherently involve copying data, copyright regulations play a pivotal role in the digital landscape, where rapid and flawless duplication is prevalent.
Nevertheless, there exists a safeguard within copyright law known as fair use. Embedded within the Copyright Act, fair use delineates certain permissible forms of copying. Due to the inability of the law to anticipate all potential scenarios, a four-factor test is incorporated to assist courts in determining the fairness of a particular copy.
However, the legal system’s application of fair use is subjective and non-uniform. Each court possesses the discretion to interpret the test independently, with no binding precedent established across different jurisdictions.
Consequently, fair use determinations are subjective and unpredictable, resembling a “vibes-based” assessment. The outcome of many copyright litigations remains uncertain, akin to a coin toss, especially when compounded by the hype, ambiguity, and financial interests associated with AI technologies.
The implications of these legal battles are profound. As discussed in the episode, the repercussions could potentially pose an existential threat to the modern AI industry, akin to the challenges faced by Napster and other file-sharing platforms in the early 2000s. Historical precedents indicate that copyright rulings during the file-sharing era led to the demise of entire enterprises and catalyzed lasting transformations in copyright legislation.
