Breaking AI on purpose: How researchers are helping make artificial intelligence safer

Skip to main content News School Logo Link MENU University of Florida For Faculty Faculty Resources Senior Communicators For Media Contact Subscribe to UF News Releases UF Facts for Media UF Administration UF Directory Hurricane Hub Breaking AI on purpose: How researchers are helping make artificial intelligence safer Harrem Monkhorst February 19, 2026 Share Nullspace steering. Red teaming. Jailbreaking the matrix.   A paper written by University of Florida Computer & Information Science & Engineering, or CISE, Professor Sumit Kumar Jha, Ph.D., contains so many science fiction terms, you’d be forgiven for thinking it’s a Hollywood script.   But Jha’s work is decidedly focused on real life, most notably strengthening the security measures built into AI tools to ensure they are safe for all to use.   “We are popping the hood, pulling on the internal wires and checking what breaks. That's how you make it safer. There's no shortcut for that.” —Sumit Kumar Jha, Ph.D., a UF professor in the Department of Computer & Information Science & Engineering As AI assistants move from novelty to infrastructure, helping write code, summarizing medical notes and answering customer questions, the biggest question isn't just what these systems can do, but what happens when they are pushed to do what they shouldn't.  “By showing exactly how these defenses break, we give AI developers the information they need to build defenses that actually hold up,” Jha said. “The public release of powerful A...