Prisoner’s Dilemma Agent

Agentic AI
Game Theory
Multi-Agent Systems
Python
Multi-agent system for answering trivia questions with LLMs.
Published

June 26, 2026

This project explored how autonomous agents behave in a repeated Prisoner’s Dilemma environment, where each agent must decide whether to cooperate or defect based on strategy, incentives, and prior interactions.

The project connects game theory, multi-agent systems, and AI decision-making.

Note

This was a group project based on a copied repository. My contribution focused on understanding the agent strategies, testing behavior, interpreting outcomes, and preparing the project for portfolio presentation.

Project Overview

The Prisoner’s Dilemma is a classic game theory problem used to study cooperation and competition. In this project, agents repeatedly interact and make decisions based on programmed strategies or learned behavior.

The simulation helps demonstrate how simple decision rules can lead to complex strategic patterns over time.

What I Worked On

  • Tested agent behavior across repeated game rounds
  • Compared cooperation and defection patterns
  • Interpreted strategy outcomes using a game theory lens
  • Reviewed the codebase and documented the agent workflow
  • Prepared the project for public GitHub presentation

Technical Focus

  • Python
  • Multi-Agent System
  • Game Theory
  • Agentic AI
  • Simulation
  • Strategic Behaviour
  • What This Project Demonstrates

    This project shows my interest in building AI systems that do more than generate text. It demonstrates how agents can be used to model strategic behavior, decision-making, cooperation, and competition in simulated environments.

    Repository

    View GitHub Repository ↗

    ← Return to all projects