Sorbonne University for a New Deal on Artificial Intelligence

Project Information

SOUND.AI

Grant agreement ID: 101081674

DOI

10.3030/101081674

EC signature date 10 October 2022

Start date 1 January 2023

End date 31 December 2027

Funded under

• Marie Skłodowska-Curie Actions (MSCA)

Total cost

No data

EU contribution

€ 3 024 000,00

Coordinated by

SORBONNE UNIVERSITE
France

Periodic Reporting for period 1 - SOUND.AI (Sorbonne University for a New Deal on Artificial Intelligence)

Reporting period: 2023-01-01 to 2024-12-31

Summary of the context and overall objectives of the project

The SOUND.AI program (Sorbonne University for a New Deal on Artificial Intelligence) is a prestigious initiative led by the Sorbonne Center for Artificial Intelligence (SCAI) and the Sorbonne University Alliance. It advances excellence in doctoral training, international mobility, and career development, fostering innovation within one of the world’s most dynamic AI ecosystems.

Launched in 2023, SOUND.AI recruited 26 exceptional PhD candidates through a highly selective process. The program is designed to:

✔ Develop a cutting-edge AI-focused PhD program, preparing students for the future of AI research.
✔ Enhance interdisciplinary learning, integrating fundamental and applied research in “societies, languages, and cultures”, “global health” and “sustainable resources”.
✔ Modernize doctoral education, strengthening PhD training across Sorbonne University and its partners.
✔ Encourage critical thinking and innovation, pushing students to expand their perspectives through workshops, summer schools, and digital training.
✔ Strengthen research collaborations, fostering connections with academic and industry leaders to open new research opportunities.

Backed by the Sorbonne University Alliance—including Sorbonne University, the National Museum of Natural History (MNHN), and the University of Technology of Compiègne (UTC)—the program integrates cutting-edge research, pedagogical innovation, and multidisciplinary collaboration to transform PhD training.

A strong network of associated partners further enriches the program, with organizations such as APHP (health innovation), BnF (cultural data), IRCAM (artistic creation), IFPEN (renewable energy), LNE (AI systems certification), and the City of Paris (public outreach) providing specialized training and real-world applications.

The program also fosters global collaborations with leading institutions such as UCL in Europe and MILA and OBVIA in North America, enabling co-supervision, international mobility, and access to world-class training opportunities. Simultaneously, industrial partners play a pivotal role in bridging research and practice, offering fellows exposure to diverse professional environments across corporates like EDF or Valeo, SMEs like CRITEO, Naver Labs, and innovative startups (Datacraft, Hugging Face, OpenClassrooms).

With its innovative training approach, global partnerships, and strong academic-industry ties, SOUND.AI is shaping the next generation of AI researchers and innovators. By reinforcing Sorbonne University’s leadership in AI research and equipping students with advanced skills, the program ensures that its fellows are ready to tackle global challenges and lead AI-driven advancements.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The project successfully implemented a robust and transparent recruitment process, attracting highly qualified candidates and promoting excellence in interdisciplinary research. A dedicated website and portal https://soundai.sorbonne-universite.fr were established to manage applications for both PhD subjects and candidates, with dedicated and secured satellite websites, https://supervisors.soundai.sorbonne-universite.fr/dl/home and https://applicants.soundai.sorbonne-universite.fr/dl/home.

A fair and transparent 2-step selection process was designed, focusing on the excellence of applicants and the quality of projects, with particular attention to interdisciplinary, international, and intersectoral dimensions. After a pre-selection based on paper evaluation, and a selection after a 15mn audition, this resulted in 239 eligible applications across both calls, leading to the recruitment of 26 fellows—8 from EU countries and 18 from non-EU countries.

The fellows cover a wide range of scientific topics, with all three faculties of Sorbonne Université actively involved. A perfect gender balance was achieved, with 13 female and 13 male fellows, without the need for specific affirmative actions.

All 26 fellows have officially started their PhDs, supported by dedicated supervisors who have shown strong commitment to both the project and the management team.

Two Induction Weeks were organised, the first in January 2024, and the second in January 2025, to welcome the fellows and provide them with scientific, technical and administrative introductions. These two weeks were mandatory for the recruited fellows but also open to doctoral students from Sorbonne Université. Several sessions specifically catered to the SOUND.AI fellow were included.
The first day was focused on the understanding of the status of a PhD student at Sorbonne Université. The SOUND.AI Program Director and Program Manager explained the COFUND programme, its specificities and functioning and key-personnel in the morning. In the afternoon, the College Doctoral of Sorbonne Université provided detailed information on life as a PhD student and their obligations.
During the week, expert scientists gave a dozen seminars and practicals about applications of AI in Science, Engineering, Medicine, Language and Humanities, from basics to more advanced topics, among them a whole day dedicated to Ethics, Scientific Integrity and Epistemology. These sessions correspond to 6 training modules and a part of the network-wide training program proposed in SOUND.AI and scientific/technical knowledge and skills for the Personal Career and Development Plan each fellow has to fill up for the end of their PhD.

Two channels are used for communication between SOUND.AI staff and fellows. For dedicated messages, email is preferred, using the most relevant email address provided by the fellow, with an institutional email address as much as possible. For generic messages, such as event announcements, reminders about documents to provide or to fill up for reporting, a SOUND.AI Mattermost channel has been open since Oct 2023. For document exchange and daily follow-up of expenses, research and training activities, or travel, each fellow has a personal and secure space on the SOUND.AI website https://soundai-laureates.glide.page/dl.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Not applicable yet

SOUND.AI logo