PhD Position on Dissecting Algorithmic Collusion

About the role

Four-year fully funded PhD at the University of Twente investigating how learning algorithms can collude (consciously or unintentionally) in market settings, and what mechanisms detect or prevent that behaviour. Sits between computer science, economics, and competition policy. Deadline 14 June 2026.

Responsibilities

• Build multi-agent reinforcement learning simulations of pricing and bidding markets.
• Identify conditions under which agents settle on collusive equilibria and design auditable signals.
• Publish in NeurIPS, ICML, AAAI, and economics-and-computation venues (EC, WINE).
• Engage with Dutch and EU competition regulators on policy implications.

Requirements

• Master’s degree in computer science, applied mathematics, economics, or a related field with strong computational component.
• Solid grounding in reinforcement learning or game theory.
• Python proficiency; familiarity with at least one RL framework.
• Fluent English. Dutch is not required.

Nice to have

• Prior work in multi-agent ML, auction theory, or industrial organisation.
• Familiarity with competition law or regulatory contexts.
• A first-author paper or preprint.

How to apply

Open the apply link and submit CV, transcripts, motivation letter, Master’s thesis, and two reference contacts via the University of Twente application portal before 14 June 2026.