Generative AI and Copyright: Paris Meeting to Examine Creators’ Income and Platform Recommendations

Artists, technology companies, copyright professionals and researchers will meet at UNESCO headquarters in Paris on 27 October for an international congress on human creativity and the diversity of cultural expressions in the digital environment. The debate over generative AI and copyright will extend from how works enter technological systems to creators’ earnings and who determines their visibility once they are published.

The official programme includes three discussions: human creation and synthetic content; AI systems that respect rights; and the discoverability of diverse cultural content. Together, they trace a production and distribution chain, from making and using works to placing them on platforms and finding audiences and income.

Digital income grows, without guaranteeing a stable livelihood

The congress comes as cultural sectors have become more dependent on digital channels. UNESCO’s 2026 Re|Shaping Policies for Creativity report puts digital revenue at 35% of creators’ income, up from 17% in 2018. Yet it also stresses that digital transformation has not secured a stable livelihood for most creators, with income instability and precarious employment remaining prominent concerns.

As income moves online, changes to platform rules have a more direct bearing on creative careers. Search visibility, recommendations and revenue allocation can determine whether a work has a chance to recover its costs. Generative AI’s expansion of the content supply brings those existing problems into a new technological setting.

A study commissioned by the International Confederation of Societies of Authors and Composers, CISAC, and conducted by PMP Strategy in 2024 estimated that, under an unchanged regulatory framework, 24% of music creators’ revenue and 21% of audiovisual creators’ revenue could be at risk from generative AI by 2028. These are modelled projections, not income declines already recorded in 2026. As an organisation representing creators’ rights, CISAC is also a stakeholder in the debate.

The study examines substitution: when some clients can buy or generate synthetic content, commissions and payments that previously went to human creators may shrink. The effects vary between occupations. An industry-level forecast cannot be read as the personal loss of every musician, screenwriter or director.

Consent and payment are separate questions

The second Paris session will consider consent, remuneration and the use of copyrighted works throughout an AI system’s lifecycle. Its scope reaches beyond the final output to how systems are developed and operated.

Whether a work is used, the conditions governing that use and whether payment follows are related but distinct questions. Technical capability cannot settle them. Creators need sufficient information to make choices; companies seeking licensing partnerships need workable arrangements defining uses, rights and payments.

Nor does every use of AI in creative work fall into the same category. The opening session will distinguish technology that assists human creation from automated content generation. How much a tool contributes, and which judgments and choices remain with a creator, shape the understanding of the creative process.

More available content does not mean more voices are heard

The final session addresses discoverability: a work may be online without its potential audience being able to find it. This matters particularly for local content, work in different languages and authors who have yet to build a reputation.

UNESCO’s 2026 report identifies the concentration of streaming platforms and opaque content curation as disadvantages for less well-known creators. Increasing the number of available works is not the same measure as broadening what audiences actually encounter.

For cultural institutions, digitising a work or uploading it to a platform is therefore insufficient evidence that distribution has succeeded. Further assessment needs to consider whether the work appears in searches and recommendations, reaches people beyond its existing audience and generates sustainable income.

Technology may also help address these questions. The congress plans to present practical experiences supporting cultural content’s discoverability and discuss how technological development can serve diverse cultural expressions. The comparison extends beyond production speed to which works gain new opportunities and which voices still struggle to reach the public.

The 27 October meeting will offer a webcast and interpretation in English and French. According to its concept note, the discussions will inform UNESCO’s subsequent action, with a report to the eleventh Conference of Parties to the Convention on the Protection and Promotion of the Diversity of Cultural Expressions in March 2027.

Sources: UNESCO congress programme and concept note; 2026 Re|Shaping Policies for Creativity report; CISAC/PMP Strategy’s 2024 study of generative AI’s economic impact.

Image: AI-generated editorial illustration.