SCAI: Meeting Scientific Challenges in the Age of AI

Could you remind us of SCAI’s mission?

Xavier Fresquet
SCAI’s ambition is to bring together researchers, students, and companies from a wide range of disciplines around shared projects, develop international hosting capacity, support early-career research, and promote a genuine scientific culture of AI through dedicated training programs and events. The center helps create an environment that fosters innovation by placing interdisciplinarity and scientific excellence at the core of its mission.

How do you bring together stakeholders as multidisciplinary as cultural centers, research institutes, schools, and universities? How are their exchanges and collaborations organized?

Xavier Fresquet

Xavier Fresquet
Multidisciplinarity lies at the heart of SCAI’s identity, which works to create dialogue between the exact sciences, the humanities, artistic creation, and engineering. This dynamic has grown even stronger with the launch of the PostGenAI@Paris cluster, a unique consortium bringing together academic institutions, national research organizations, more than 60 companies, as well as cultural institutions such as IRCAM and the Bibliothèque nationale de France.

Collaborations are structured through 21 Collaborative Acceleration Programs (PACs), each of which systematically brings together several disciplines around a shared scientific challenge, as well as through a cross-disciplinary program designed to circulate expertise and continuously adapt research themes. The cluster operates through collective governance, making it possible to connect lawyers, mathematicians, acousticians, engineers, economists, artists, and heritage experts through shared projects.

The cluster represents a new phase in the evolution of AI: one in which models no longer simply generate content, but are also capable of evaluating, understanding, reasoning, and interacting in context-aware ways. Its missions are organized around three main priorities: developing breakthrough technologies in generative AI, placing health and biodiversity at the center of a desirable future, and strengthening social resilience through AI, notably by supporting the arts, digital humanities, and governance.

Today, PostGenAI@Paris positions itself as a true international hub for post-generative AI.

In your view, what notable SCAI projects emerged in 2025?

Xavier Fresquet
The past year marked an important milestone with the rapid growth of the PostGenAI@Paris cluster. Among the major projects are the European MSCA COFUND program “SOUND.AI,” dedicated to 30 doctoral students, the educational developments of the AISorb project — launched through the “Skills and Professions of the Future” (CMA) call for expressions of interest — which expand AI education at both undergraduate and master’s levels, and the contribution of the Campus of Excellence for Careers and Qualifications in AI, aimed at extending AI education into high schools across the Île-de-France region.

The year also saw the launch of new creative and transdisciplinary initiatives, including [G.R.@.B.U.G.E], the “Sound Research Group and Other Utopian, Brilliant, and Ephemeral DIY Experiments,” imagined by Nicolas Obin and supported by Sorbonne Université, IRCAM, SCAI, and the Collegium Musicæ. This low-tech, joyful, and openly experimental initiative explores sound-making, musical tinkering, and unusual devices — including sonic interactions with mushrooms — and reflects the vitality of collaboration between art, technology, and research within SCAI.

Nathalie Drach-Temam, présidente de Sorbonne Université lors de l'inauguration du cluster PostGenAI@Paris, avril 2025

IRCAM has been collaborating with SCAI since 2020. How do you see the integration of its disciplines into SCAI’s program?

Xavier Fresquet
The collaboration with IRCAM is a key pillar in the development of AI for sound, music, and creative practice. It helps open AI research to fields where perceptual, aesthetic, and technical demands are especially high.

Several PACs structure this integration: AI-MADE, which aims to develop innovative generative models for professional audio content generation, and TWINNING, which addresses the broader issue of autonomous vehicles. Research on sound perception and cognition, as well as Creative AI training modules designed for artists and students, further strengthen this collaboration.

IRCAM also plays a leading role in the PAC dedicated to digital humanities, and in broader reflections on the place of AI in contemporary cultural practices.

AI is taking on an increasingly important role in the cultural and creative industries. What challenges must the stakeholders involved in SCAI address? And personally, how do you view this development?

Xavier Fresquet
The rise of generative AI in the cultural industries opens up immense possibilities, but it also calls for increased vigilance. The first challenges concern ethics and governance: distinguishing between the real and the virtual, protecting data, preventing misinformation, and respecting copyright

There are also technical challenges, including the need to develop models capable of producing reliable, traceable, and non-harmful content.

Finally, there are artistic challenges: supporting creators, inventing new tools, and ensuring that AI remains a driver of experimentation and freedom, rather than standardization.

Personally, I remain optimistic, but demanding. AI can amplify human intelligence and creativity, provided that it is developed within a robust ethical framework and through constant dialogue with artists, researchers, and citizens alike.

Campus Pierre et Marie Curie - Photo Pierre Kitmacher © Sorbonne Université

What are SCAI’s major projects for addressing these rapid transformations shaping our societies?

Xavier Fresquet
The priorities for 2026–2030 are numerous. They include the development of a Continuing Education Office, offering certificates and professional training programs focused on AI ethics, healthcare applications, and AI for climate-related issues, as well as the scientific renewal of the PACs to integrate new emerging fields.

One major focus will be the development of the broad field of AI & Mathematics, which will fully expand starting in 2026. This field, which is fundamental to the theoretical understanding of generative models and reasoning systems, will also have implications for musical creation, sound analysis, and acoustic modeling by strengthening ties between mathematics, IRCAM, the Collegium Musicæ, and experimental projects such as G.R.@.B.U.G.E.

Finally, the opening of the Cité de l’innovation (Innovation City) at Sorbonne Université and the organization of major international scientific events will help position Paris as a global center for post-generative AI, capable of supporting technological developments while keeping people at the center.