Arslan Bisharat
Ph.D. Student
Teaching & Research Assistant
Arslan Bisharat studies whether AI systems can be trusted when it matters. A model that scores well on a benchmark is not the same as a model that holds up in the real world, and that gap is what his research is about.
His work develops trustworthy AI across three fronts: online safety, formal reasoning, and security. He studies how harmful behavior such as cyberbullying spreads across social platforms, because detection models are only trustworthy if they reflect how that behavior actually moves between people and communities. He builds benchmarks that test whether large language models can reason over formal specifications, because a claim about machine reasoning deserves verification rather than assumption. And he examines how federated learning systems fail under adversarial pressure, because distributed AI carries security risks that are easy to overlook until they cause harm.
The thread running through all of it is evaluation. Trust in an AI system has to be earned through evidence, and Arslan builds the tools that reveal where these systems break down before the consequences reach the people who depend on them. That is the question he keeps returning to, and the one he plans to build his career around.
He conducts this research under the supervision of Dr. Yasin N. Silva and Dr. Mohammed Abuhamad at Loyola University Chicago.
Major
Education
- BS Computer Science (Completed)
- MS Data Science (Completed)
- PhD Computer Science (In Progress)
Research Interests
AI evaluation and benchmarking, cyberbullying detection, cross-platform contagion analysis, social computing, large language model reasoning, formal verification, federated learning security, and adversarial robustness.
Specialty Area
Trustworthy AI