In brief
- UNESCO has projected that generative AI outputs could drive revenue losses of 24% for music creators and 21% for audiovisual creators by 2028.
- The report found digital revenues now make up 35% of creators’ income, up from 17% in 2018.
- Copyright lawsuits against OpenAI, Google, Meta, and Anthropic add to mounting legal issues over AI training practices.
AI is poised to take a measurable bite out of the world’s creative economy, and the legal tools meant to protect artists are already showing their seams.
UNESCO’s latest Re|Shaping Policies for Creativity report, a global monitoring study drawing on data from more than 120 countries, forecasts that generative AI outputs could trigger global revenue losses of up to 24% for music creators and 21% for audiovisual creators by 2028, as AI-generated content expands and competes with human-made work.
Ishita Sharma, managing partner at Fathom Legal, told Decrypt the projections “significantly strengthen the normative case for recalibrating copyright and neighboring rights frameworks,” adding that the debate has moved from abstract innovation to “distributive imbalance” when AI systems extract value from protected works “at scale without proportionate compensation.”
That matters because creators are increasingly reliant on digital channels that are both lucrative and unstable.
UNESCO found digital revenues now represent 35% of creators’ income, up from 17% in 2018, even as public funding for culture remains below 0.6% of GDP globally, a change that has widened exposure to IP infringement while platform power and opaque curation systems push lesser-known artists to the margins.
Sharma noted that “existing fair use and fair dealing doctrines appear increasingly strained in the context of AI training and output replication,” saying the doctrines were built to assess “human, case-specific transformative uses—not the wholesale ingestion of vast copyrighted corpora for commercial model development.”
Even if training is framed as transformative, she said, deploying outputs that substitute for original works “complicates the analysis,” often leaving uncertainty that advantages large technology firms.
The UNESCO report also flagged structural gaps that could amplify those losses, including a digital skills divide, 67% in developed countries versus 28% in developing countries, and limited national capacity to measure digital cultural consumption.
For creators whose style or voice is replicated, Sharma said remedies remain “fragmented and imperfect,” because copyright protects specific expression rather than “style,” while publicity and false-endorsement claims are uneven across jurisdictions, fueling calls for clearer remuneration-focused protections.
The question of who benefits from AI’s growth is being actively contested in legal proceedings.
OpenAI and other AI firms are facing mounting copyright lawsuits from authors and publishers alleging unlicensed book downloads and training practices, with a New York judge recently allowing key infringement claims to proceed in consolidated litigation.
Publishers have joined litigation against Google, alleging it copied books without permission to train its Gemini AI, while judges have granted Meta and Anthropic partial fair use victories in similar book-training cases.
In Hollywood, more than 500 writers, actors, and technologists have backed the Creators Coalition on AI, calling for enforceable standards governing how AI systems are trained and how creators are compensated.
At the same time, parts of the tech sector are investing in adaptation.
Google.org announced a $2 million investment in the Sundance Institute to train more than 100,000 artists in foundational AI skills, positioning AI literacy as a baseline creative competency.
Daily Debrief Newsletter
Start every day with the top news stories right now, plus original features, a podcast, videos and more.
