PeopleofSuperintelligence

Researcher

Gabrielle Berrada

Gabrielle Berrada is an AI researcher whose work at Mistral AI spans reasoning, speech, model evaluation and multimodal safety. She studied for an engineering diploma at École Nationale Supérieure des Mines de Paris and Université Paris Sciences et Lettres, studied at Oxford in 2023–2024, and worked in research at ETH Zurich and machine learning at Aqemia. In 2025, she coauthored Magistral, Mistral's first reasoning model, Voxtral audio models, and NeurIPS research on active testing for large language models. Since 2025, she is an AI Scientist at Mistral AI; her 2026 work includes Shieldstral, a 3B-parameter multimodal safety classifier.

Timeline

  1. 2021–2025

    Engineering studies in Paris

    She studies for an engineering diploma at École Nationale Supérieure des Mines de Paris and Université Paris Sciences et Lettres from 2021 to 2025. At the University of Oxford, she is an MS student from 2023 to 2024.

  2. 2022–2023

    Research work in Zurich

    At ETH Zurich, she works as a research assistant from 2022 to 2023. At Aqemia, she is a Machine Learning Engineer Intern in 2023.

  3. 2025

    Magistral reasoning model

    She coauthors Magistral in 2025, Mistral's first reasoning model, with a scalable reinforcement-learning pipeline.

  4. 2025

    Voxtral audio models

    She coauthors Voxtral in 2025; the paper presents Voxtral Mini and Voxtral Small as multimodal audio chat models.

  5. 2025

    NeurIPS work on active testing

    She coauthors the NeurIPS 2025 paper Scaling Up Active Testing to Large Language Models. It shows how cheaper surrogate models can scale label-efficient testing to LLMs and improve evaluation against random data selection.

  6. Since 2025

    AI Scientist at Mistral

    She is an AI Scientist at Mistral AI. In 2026, she coauthors Shieldstral, a 3B-parameter policy-adaptive multimodal safety classifier.

  1. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.
  2. The United States employs an estimated 53% of people in our data with a known employment countryOur estimates of where AI careers are concentrated now and how countries’ shares of dated positions changed over time.
  3. 46% of living people in our data with known birth and employment countries work outside their birth countryOur estimate of cross-border careers in AI, from country of birth and the country of each current organization.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionOur estimates of where the people shaping AI studied and where their careers led, from education records in cited biographies.

Sources and credits

Photo: We have no published photo of Gabrielle Berrada yet. Send us a free-licensed photo.

  1. Gabrielle Berrada — Lacuna lacuna.tiptreesystems.com. Checked 19 October 2026.
  2. Magistral arxiv.org. Checked 19 October 2026.
  3. Voxtral arxiv.org. Checked 19 October 2026.
  4. Scaling Up Active Testing to Large Language Models arxiv.org. Checked 19 October 2026.
  5. Scaling Up Active Testing to Large Language Models proceedings.nips.cc. Checked 19 October 2026.
  6. Shieldstral arxiv.org. Checked 19 October 2026.