BJBrian Jalaian

Associate Professor · AI Research Leader

What post-training changes in AI agents, and when we can trust them.

I direct the Artificial Super Intelligence (ASI) Lab at the University of West Florida, where I build and study agentic AI: language and vision-language agents that use tools, reason over many steps and work in teams. Every deployed agent runs on a post-trained model, fine-tuned, distilled or quantized. My research asks what those steps change in how agents behave: their reliability, uncertainty and safety, including robustness to prompt injection. My group builds multi-agent systems that know when they're unsure, and rigorous evaluation that tells a real improvement from noise. The goal is a scientific basis for deciding when people can trust an AI agent.

Brian Jalaian
Brian Jalaian, Ph.D.Pensacola, FL

Latest

All news
  1. Nov 2026

    Coming soon: Can We Trust AI?, a research podcast

    A short preview of Can We Trust AI?, a narrow research podcast built around the questions my lab works on. The first episodes are coming in the next few days.

  2. Nov 2026

    Rain blurs the camera. Radar still sees.

    A highway in heavy rain, staged end to end by our simulation pipeline: the camera view on top, and the radar and lidar view of the same moment below.

  3. Nov 2026

    Safety in Numbers, in AI Magazine

    "Safety in Numbers: Tackling the deliberative, performative, and infrastructural roots of multi-agent misalignment", led by Jeff Bradshaw at IHMC with 22 co-authors including me, is out in AI Magazine (open access).

Follow the work onYouTubeXLinkedInInstagramRSS

Work with me

I'm building trustworthy agentic AI, and looking for the right collaborators, students, and problems.