In a striking development that highlights the ongoing tensions between authors and technology companies, a coalition of writers, led by John Carreyrou—best known for exposing the Theranos scandal and authoring "Bad Blood"—is initiating legal action against six prominent AI firms: Anthropic, Google, OpenAI, Meta, xAI, and Perplexity. The crux of their lawsuit revolves around allegations that these companies have utilized unauthorized copies of their books to train their artificial intelligence models.
This situation may sound familiar; previously, another group of authors launched a class action lawsuit against Anthropic, alleging similar violations of copyright laws. In that earlier case, a judge determined that while it was permissible for Anthropic and comparable AI enterprises to leverage pirated book copies for training purposes, the act of pirating those books initially was not lawful.
As a result of that settlement, eligible authors are set to receive approximately $3,000 from the $1.5 billion payout from Anthropic. However, many authors expressed dissatisfaction with this resolution. They argue that it fails to hold AI companies accountable for their actions of using stolen intellectual property to develop models that generate substantial revenue—billions, in fact. Recent data reveals that popular applications like ChatGPT's mobile platform have reached a staggering milestone of $3 billion in consumer spending.
The new lawsuit underscores the sentiment among the plaintiffs, who claim that the previous settlement appeared to benefit the AI companies rather than the creators of the content they exploited. The lawsuit articulates a crucial concern:
"Companies developing large language models (LLMs) should not be allowed to easily dismiss countless high-value claims at minimal costs, thereby sidestepping the true implications of their extensive and willful copyright infringements."
This situation certainly raises important questions about the ethics of AI training practices and the protection of intellectual property rights. As this debate unfolds, what do you think? Are AI companies truly accountable for their use of copyrighted materials, or does the existing legal framework provide them with too much leeway? We invite you to share your thoughts and opinions in the comments below!
