This course studies how artificial intelligence can be used as an oracle in multiagent economic systems. Students learn to specify the inputs and outputs of an AI oracle, use its predictions or recommendations in platform and economic decisions, and analyse how agents respond strategically to oracle-assisted rules. Through modelling exercises, computational labs, and small-scale simulations, students formulate economic objectives as explicit and testable requirements and verify whether system outcomes satisfy them.

For a student to pass the course, at least 30% of the maximum mark for the continuous assessment and 30% of the maximum mark for the examination must be obtained.

Applied Project

The project examines a selected AI-mediated multiagent economic system. Possible topics include:

• an advertising auction using AI-predicted conversion probabilities;


• a recommender system with exposure-fairness requirements;


• a matching market using AI-generated compatibility scores;


• auto-bidding agents in sponsored search;


• marketplace ranking with strategic sellers; or


• multiagent negotiation or bargaining using large language models.

The report must include a problem formulation, a formal model, a specification of the AI oracle's inputs and outputs, an economic or platform decision rule, modelling or simulation results, explicit verification requirements and results, economic interpretation, and limitations. Implementation of a complex AI model from first principles is not required.