What is ATLANTIS?
The project
ATLANTIS is a five-year research project on computational antitrust, led by Thibault Schrepel, based at Vrije Universiteit Amsterdam and funded by the European Research Council (Consolidator Grant, two millions €). Antitrust agencies now rely on computational tools to enforce the law. We build the legal regime that lets them do so accurately, fairly, and sustainably.
The premise
Most antitrust agencies now use computational tools to detect infringements, analyze practices, and monitor remedies. These tools include data mining, screening techniques, machine learning, natural language processing, and network analysis. They have become essential to enforcement in fast-moving digital markets.
Their adoption has outpaced the law that should govern them. Agencies struggle to collect reliable data, AI systems carry hidden biases, and courts overturn decisions that cut corners. ATLANTIS supplies the legal framework that keeps computational antitrust effective and in check.
Method
We work with a global network of antitrust agencies to document how they use computational tools, what data they need, and where the law constrains them.
From there we build instruments. Datasets, tools, and legal proposals that regulators and scholars can use, tested against how enforcement actually works.
Audience
ATLANTIS addresses several audiences. For agencies, the project will produce guidelines, policy briefs, workshops, and training sessions. For scholars, it will publish articles, datasets where appropriate, calls for papers, and conference materials.
For the public, it will provide short explainers, updates, recordings, and findings written in accessible language.
Info box
Funded by: European Research Council, grant agreement 101228709
Host institution: Vrije Universiteit Amsterdam, Faculty of Law
Project duration: 2026-2031
Grant value: 2 millions euros
Principal investigator: Thibault Schrepel
Research home: Amsterdam Law & Technology Institute
Work streams
Accurate: Data collection that respects the rights of defense
Antitrust agencies must collect company data to assess practices, often within legal regimes built for a pre-digital world. Project A asks how they can do so effectively while safeguarding fundamental rights.
Fair: Procedural fairness and AI bias in agency decisions
AI systems trained on past enforcement carry biases that distort priorities and outcomes. Project B brings the dedicated literature on transparency and explainability into antitrust to restore procedural fairness.
Sustainable: Institutional frameworks that keep antitrust working
Agencies are experimenting with different institutional setups for computational tools, with little research on what works. Project C draws on institutional economics to consolidate the evidence and build a durable framework.
Contact
Dr. Thibault Schrepel, LL.M.