Project Description
This project is a direct continuation of our Phase 1 research, which mapped Russian disinfolklore to its original sources. The overarching objective is to develop a semi-autonomous, multi-agent AI framework for counter-disinfolklore. The methodology involves three students, each building a specialized AI agent inspired by Sefton Delmer's techniques: a Persona/Inoculation Agent, a Demoralization/Chaos Agent, and a Call-to-Action Agent. This system will be designed with a workflow tool like n8n to generate strategic counter-narratives based on a fine-tuned General Language Model. We will collect data from online communities and ultra-right websites. The research findings will be integrated into the curriculum of courses on Information Security and Responsible AI. In collaboration with experts, we aim to produce a peer-reviewed publication and a functional prototype of the counter-narrative system, bridging innovative AI application with vital information literacy and entrepreneurship.
Tasks and Responsibilites
The student will work with Simon Potapov on the task to build with n8n tool an AI agent to recognize and analyze disinfolklore narratives originating directly from Russian sources.
Desired Qualifications
None Listed