We aim to use Reinforcement Learning to allow an agent to be able to conduct penetration tests without a human operator. We aim to leverage the artificial intelligence and machine learning methods for intelligence analysis to combat Human Trafficking. We develop machine learning models to collect data from sources such as social media, news websites, forums, and the darknet to discover and identify indicators of threat. We study and develop reliable and scalable machine learning techniques to identify vulnerabilities, detect malicious behavior, and prevent attacks. The medical device industry is leaping forward by relying on electronics to improve life-saving medical devices’ safety and performance. Adversary Engagement Ontology (AEO) is a subset of the Unified Cyber Ontology that aims to define and standardize the representation of This project investigates whether Motor Imagery (MI) BCIs are vulnerable to Adversarial Stimuli, which are minor sensory perturbations. Multi-Agent Systems (MAS) is the study of multi-agent interactions in a shared environment. Communication for cooperation is a fundamental construct for sharing information in partially observable environmentsAbout Us
The Secure and Assured Intelligent Learning (SAIL) lab works towards laying concrete foundations for the safety and security of intelligent machines with both theoretical and engineering perspectives. Our research aspires to develop comprehensive models, metrics, frameworks, and tools for analysis, implementation, and mitigation of deleterious behaviors in AI systems.
In The Media
Latest Events
News
Automated Penetration Testing using Reinforcement Learning
Combating Human Trafficking via Automated Intelligence Collection, Validation, and Fusion
Automated Collection and Analysis of Open-Source Cyber Threat Intelligence
Machine Learning for Cybersecurity
FAULT DETECTION AND PROGNOSIS IN MEDICAL DEVICES
ADVERSARY ENGAGEMENT ONTOLOGY
ADVERSARIAL ATTACK ON EEG-BASED BCI DEVICES
DECEPTION IN MULTI-AGENT SYSTEMS