Behavioral Response Analysis, Integration Network and Simulation
BRAINS is an advanced AI-driven simulator that recreates population behavior. It is a platform that allows organizations to predict how specific groups of people will perceive and react to information or initiatives before they are launched. It solves the critical problem of public resistance and misinformation, which cause the failure of technically sound projects, like: green infrastructure or smart cities initiatives; due to a lack of societal acceptance. Currently, the project has been applied to different use cases for sustainability and for strategic decision support capable of simulating complex real-world scenarios.
Target Customers and Buyer Groups
The simulator is particularly useful for high-stakes decision-makers across several sectors considering Government and Urban Planners, Industrial Leaders, Political and Strategic Communicators and Marketing Professionals. It creates measurable value by bridging the gap between technical implementation and public endorsement improving media campaign success rates, decision making, and risk mitigation.
Brains is a proactive and predictive tool that uses Stochastic Simulation to model future scenarios instead of just analyzing historical data. Using Human Behavior Modeling it replicates population characteristics and reactions considering cognitive biases (confirmation bias, bandwagon effect).
It integrates Generative AI to adjust and improve the users communication with the population to achieve the desired effect. The tool suggest to the decision-maker how to adjust the information campaign according to the desired channel of communication (Broadcasting, Press and Social Media).
Scientific and Methodological Backbone
The platform relies on the Strategic Engineering approach, which provides a rigorous framework combining:
- Agent-Based Modeling (ABM): Utilizing Intelligent Agents (IA) to represent diverse population segments based on scraped demographic data (age, education, politics, etc.).
- Human Behavior Models (HBM): Advanced models that reproduce psychological attributes and social network ties.
- Large Language Models (LLMs): Integrated into the simulation to perform Sentiment Analysis and create authentic content.
- Information Diffusion Theory: Modeling how information spreads through “cascades” and “echo chambers” within social networks.
This platform has been applied to two use case scenarios:
Sustainability and Behavioral Analysis: This version, known as BRAINS2, was created to address sustainability debates among different parties and to explore the use of AI for cognitive actions and propaganda in that field. It uses generative AI to create realistic communication content and performs sentiment analysis to update population parameters based on how people perceive and react to information. The goal is to simulate the cognitive impacts that media interactions have on a specific population.
Decision Support and Strategic Success: The second version, BRAINS3, is an advanced AI solution developed to analyze and counter propaganda and cognitive campaigns across multiple platforms such as TikTok, Instagram, X (formerly Twitter), and TV. It is applied to scenarios like political campaigns, international disputes, and neuromarketing. This version can self-generate messages or re-elaborate manual content to help a user succeed against opponent attacks on specific subjects or population sectors. It serves as a tool for both real-time decision support and training for managing complex opponent campaigns.
Academic Validation: The methodology has been peer-reviewed and presented at major conferences, such as the SESDE 2024 International Workshop.
Bruzzone, A. G., Giovannetti, A., Cirillo, L., Sina, X., Ghisi, F., & Ferrero, S. G. (2024). Innovative framework based on simulation and generative AI for enhance decision making in green projects. In Proceedings of the 12th International Workshop on Simulation for Energy, Sustainable Development & Environment (SESDE 2024),. https://doi.org/10.46354/i3m.2024.sesde.012