AI to slash music and audiovisual industry revenues by over 20% by 2028, report warns

Culture has long been aware of the threat posed by machines. It has been imagined in countless films, novels, and video games. Now, however, thousands of artists believe that the dystopia is encroaching on their real lives. The threat does not come from robots with machine guns — but rather artificial intelligence (AI) programs capable of stealing their works, and eventually their livelihoods.

Specifically, the music and audiovisual sectors are at risk of losing, respectively, a quarter and a fifth of their global revenue by 2028, according to the first study on the potential economic impact of generative AI on the arts. If left unregulated, AI’s advance could have severe consequences, according to the study, commissioned by CISAC, the global organization that unites major copyright management entities. The research — which provides concrete figures on the economic impact of AI on the arts — aims to spark debate about the future of the sector and, more importantly, to find solutions.

Until now, many creators have responded to this latest and most powerful technological revolution with a mix of prophetic warnings and complaints, both publicly and in court. The study highlights legal cases such as the lawsuit filed by artists Sarah Andersen, Kelly McKernan, and Karla Ortiz against Stability AI and Midjourney, as well as lawsuits filed by several major music companies against Suno and Udio.

On the other hand, AI developers and their supporters view these complaints with skepticism, dismissing them as exaggerated, alarmist, and biased, while overlooking the potential benefits these new tools offer artists. The CISAC report seeks to quantify the damage the supposed “AI apocalypse” could bring in the next five years, while highlighting the enormous profits reaped by the developers behind these technologies.

The study, conducted by PMP Strategy, a strategic management consultancy, predicts that the income of developers of generative AI — models that can “learn complex patterns to generate new content” — will soar at nearly the same pace as the cultural sector’s revenue. The report renews accusations from creators who claim that AI programs have trained themselves by consuming millions of works protected by copyright for free and without permission. First, the machines learn the style of a filmmaker, cartoonist, or musician. Then, they emulate these styles with increasing accuracy, competing directly with the creators for a share of the market.

One only needs to look at the evolution of videos produced by Sora, images created with Midjourney V6, or songs composed by Suno to understand the rapid growth of these technologies. “The later a capability started to be implemented, the faster it reached human-level performance,” the report states. For example, it took 19 years for machines to recognize the speech of children under 12 months old, but now AI can detect, understand, and use language with proficiency.

The report even speaks of the “cannibalization” of creative catalogs, estimating, for example, that music generated by computer programs could capture 20% of the revenue from streaming platforms. This prediction is also based on the growing trend of “passive” viewing or listening, which provides more space for content that may not be outstanding but serves as background entertainment or a simple way to pass the time. The increasing influence of AI-powered algorithms only exacerbates this situation. As the report points out, “75% of Netflix viewership is driven by its recommendation engine,” and Spotify now includes and can recommend AI-generated music in its playlists, according to the report.

CISAC, as a key stakeholder, represents five million artists, their works, and their concerns. This may explain why the report dedicates the first 116 pages to outlining its methodology in detail. It includes interviews with over 50 stakeholders, including creators, producers, publishers, distributors, technology companies, and AI institutions. Among those interviewed are major players like Sony, Warner, SGAE, Deezer, the European Filmmakers Federation (Fera), and the European Parliament, as well as companies such as Ask Mona, Kyutai, and Fairly Trained — all of which approach AI from a copyright-friendly perspective. Google and Microsoft were also interviewed. However, while OpenAI, the company that has garnered the most headlines, fascination, and fear in recent times, is frequently referenced in the report, it was not directly interviewed.

The study is also based on market data, AI reports, copyright laws, and eight workshop sessions between industry members and CISAC. With this comprehensive approach, the report aims to answer three critical questions for 2028: What role will AI-generated works play in the audiovisual and music markets? How much revenue will developers of these technologies generate? And what will the financial impact be for creators?

The report predicts that AI-generated music will generate revenues of $16 billion in 2028, up from $1 billion today. Similarly, the audiovisual sector could see a jump to $48 billion in five years, compared to $2 billion now. The estimated losses for artists in both sectors are projected to exceed $4 billion each, according to the document. However, there are nuances: dubbing artists and translators are likely to be more affected in the short term than directors or screenwriters.

The report also outlines some of the most plausible scenarios: AI-generated music replacing background music in shops, public spaces, or soundtracks, dominating content shared on social media, or being used in low-budget video games, films, and series to reduce production costs. In the audiovisual sector, AI could take over tasks such as creating cartoons and videos for children’s catalogs or advertisements, as well as reducing funding for scriptwriting and directing by outsourcing these tasks to computer programs.

In fact, some of this is already happening. “The report confirms that the cultural sector is in a very complicated situation if the work of authors is not respected,” says Cristina Perpiñá-Robert, general director of the SGAE, in a statement issued by CISAC. AI is no longer just a hypothetical concern; millions of workers are already using programs like ChatGPT to streamline daily tasks, and many others are engaging in debates over its advantages and potential dangers. Within the cultural sector, thousands of artists are trying to turn AI into an ally for their works. Rejecting it, therefore, no longer seems like a viable option. The report also highlights an “unprecedented” increase in private investment in generative AI in 2023.

Stopping this technological tide may seem unrealistic. However, like many other cultural organizations, CISAC urges governments to take action and guide AI toward the common good. Indeed, the regulation on AI approved by the EU earlier this year — the first of its kind in the world — makes the protection of intellectual property one of its primary concerns. In Spain, the Ministry of Culture has pledged not to award or contract works created entirely by AI.

Björn Ulvaeus, president of CISAC, states in a press release: “For creators of all kinds, from songwriters to film directors, screenwriters to film composers, AI has the power to unlock new and exciting opportunities — but we have to accept that, if badly regulated, generative AI also has the power to cause great damage to human creators, to their careers and livelihoods. Which of these two scenarios will be the outcome?”

The question was forwarded to ChatGPT, which responded: “It will depend on the decisions we make in terms of ethics, regulation, and cooperation between humans and technology.” We have been warned — AI itself has spoken.

Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition