AI’s unauthorised use of authors’ work sparks call for licences

In the ever-evolving landscape of technology and creativity, the question of authorship and fair compensation has become increasingly complex. The Authors Guild, a prominent organisation representing writers, has been at the forefront of a battle for authors’ rights. In collaboration with congressional leaders, they have been working to devise measures aimed at safeguarding the rights of authors in the realm of language-generative AI. According to Poets & Writers, one of their proposals involves the creation of a collective management organisation, responsible for granting licences for authors’ work and negotiating equitable fees paid by AI companies.

Read: Authors’ pirated books used to train Generative AI

These fees would then be distributed to authors, ensuring they receive remuneration for both past and future uses of their written content.

This is in response to language-generative AI, such as ChatGPT by OpenAI and LLaMA by Meta AI, which has brought the spotlight to the forefront of the writing community’s concerns. These AI systems are being fed millions of copyrighted books, articles, essays, and poems, raising serious questions about the use of authors’ work without proper consent and compensation.

Mary Rasenberger, CEO of the Authors Guild, emphasised the need for regulation in the AI sector, stating, “What we want to do is regulate AI, and as part of that regulation we need to make sure that writers are getting paid.”

Class-action lawsuits shake the AI industry

The urgency of this issue is underscored by the fact that several prominent authors have taken legal action against OpenAI and Meta for alleged copyright infringement. Class-action lawsuits filed since June have seen authors like Mona Awad, Michael Chabon, and comedian and memoirist Sarah Silverman seeking damages for what they perceive as the unauthorised use of their works to train language-generative AI. Lawyers representing these authors argue that AI language models “remix the copyrighted works of thousands of authors…without consent, compensation, or credit.”

Read: More authors sue OpenAI and Meta over copyright due to training

The attorneys also suggest that many of the books used in AI training likely originated from “shadow libraries,” where pirated books and publications are distributed, further complicating matters. In these class-action cases, the number of class members who can join the lawsuits is virtually unlimited, potentially leading to significant financial repercussions for AI companies.

Regarding the value assigned to authors’ works in these lawsuits, Rasenberger noted that it remains unclear how this assessment will be made. Nevertheless, plaintiffs in copyright lawsuits can potentially claim up to $150,000 per work infringed.

Revered authors like Margaret Atwood, Stephen King, George Saunders, and Zadie Smith have found their books utilised to train generative AI systems. This practice has enabled AI to replicate high-quality prose and verse, raising concerns about the ethical use of their work.

The ethical dilemma of AI replicating authors’ voices

Compensation has become particularly crucial in the current climate, with the Authors Guild reporting a 40% decline in writers’ income over the past decade. In 2022, a full-time author earned a median income of just $23,330, according to the Guild’s survey.

In response to the AI companies’ use of their work, Atwood quipped in an Atlantic essay, “Would it kill these companies to shell out the measly price of thirty-three books?” Beyond financial concerns, Atwood expressed unease at the idea that an author’s voice and creative mind could be replicated by AI.

Atwood was among the ten thousand signatories of an open letter published by the Authors Guild in June. The letter demanded that AI companies like OpenAI and Meta refrain from using authors’ works without consent and ensure proper crediting and compensation. This demand was supported by literary luminaries and advocates, including Chabon, Jonathan Franzen, Roxane Gay, and Celeste Ng.

Can you name these books and their authors?

The Authors Guild contends that AI companies cannot justify their use of authors’ works with the “fair use” legal doctrine, as they have allegedly relied on “notorious piracy websites” to access copyrighted content. According to the Guild, AI tools effectively lead to a commercial repackaging of human works, undermining the fair use argument.

Previously, authors could ascertain if ChatGPT had access to their work by prompting the chatbot to provide quotations. However, recent developments have seen ChatGPT withholding extensive quotations and denying training on “the full text of copyrighted books.” This may change as the class-action lawsuits progress, potentially leading to more transparency regarding the use of authors’ works.

Toward a fairer future: licences for authors’ works in the AI era

However, the battle for authors’ rights in the age of AI is not new for the creative body. For example, In March, the Authors Guild sought to engage with Internet Archive/Open Library, urging them to create licences for books used on Open Library. Consequently, the Internet Archive lost the lawsuit, reinforcing the idea that “scanning & lending books w/out permission or compensation is NOT fair use.”

As the fight for fair compensation continues, some voices within the writing community are advocating for licences for every single use of authors’ works, rather than just a one-time licence.

Thus this reflects the growing need to protect the intellectual property and livelihoods of authors in an increasingly digital and AI-driven world.