Mariia Vladimirova
Senior Research Scientist ·
Criteo AI Lab, Paris
AI evaluation · Fairness auditing · Causal inference · AI governance
I am a senior research scientist at Criteo AI Lab in Paris and a member of the Fairplay joint team (Inria). I work on the evaluation of AI systems: how we measure what a model does, how much the measurement depends on the protocol, and what a verdict is worth once it leaves the lab.
Most of my current work audits generative models for fairness. The same model receives opposite verdicts depending on the prompt family, decoding settings, and scorer used to test it, so I study these degrees of freedom directly and promote reproducible auditing through Fairness Cards. Earlier work built FairJob, a real-world benchmark for fairness in online advertising, and regularizers that give practitioners direct control over the fairness-utility trade-off. In causal inference, I use uplift modeling to learn and evaluate treatment policies offline for large-scale online systems, where a policy has to be judged before it can be deployed. Before joining Criteo, I did research on theoretical deep learning at Inria Grenoble 🚡🏔️. An overview of current and upcoming research directions is available on the research page.
Evaluation is also where research meets regulation. I provide technical and scientific inputs to policymakers on AI regulation, including contributions to CNIL consultations on GDPR and AI systems in the context of the EU AI Act. I also co-chair annual AI symposiums, give invited talks at universities and summer schools, and was recognized as a Rising Star in AI Ethics 2025.
@inproceedings{vladimirova2026fairness,
title={Position: Fairness Failure in Generative Models is an Evaluation Problem},
author={Vladimirova, Mariia and Franceschi, Jean-Yves and Issenhuth, Thibaut},
booktitle={International Conference on Machine Learning},
year={2026}
}@misc{daudiffret2026debias,
title={Debias Anything: Fairness with Diversity without
Supervision in Diffusion Models},
author={d'Audiffret, Th{\'e}au and Vladimirova, Mariia and
Franceschi, Jean-Yves},
year={2026},
eprint={2610.01815},
archivePrefix={arXiv},
primaryClass={cs.LG}
}@inproceedings{vladimirova2024fairjob,
title={{FairJob}: A Real-World Dataset for Fairness in Online Systems},
author={Vladimirova, Mariia and Diemert, Eustache and Pavone, Federico},
booktitle={Advances in Neural Information Processing Systems,
Datasets and Benchmarks Track},
year={2024}
}@article{arbel2026primer,
title={A Primer on {Bayesian} Neural Networks: Review and Debates},
author={Arbel, Julyan and Pitas, Konstantinos and
Vladimirova, Mariia and Fortuin, Vincent},
journal={Statistical Science},
volume={41},
number={2},
pages={316--353},
year={2026},
doi={10.1214/24-STS969}
}