Paleja earns 2026 DARPA Young Faculty Award
10-12-2026
Rohan Paleja, assistant professor of computer science at Purdue University (photo by Brian Powell).
Purdue University computer science Assistant Professor Rohan Paleja has received a Defense Advanced Research Projects Agency (DARPA) Young Faculty Award to develop new methods for ensuring teams of autonomous artificial intelligence agents operate safely and remain aligned with human goals.
Paleja’s project, “Formal Methods for Preventing Emergent Failure in AI Teams.” The project addresses a growing challenge in AI safety: While many existing approaches focus on individual AI agents, teams of interacting agents can develop unexpected behaviors that emerge from their interactions.
Through the award, Paleja will develop a framework to proactively prevent, detect, and contain failures unique to multi-agent AI systems, including covert collusion, strategic deception, and collective deviation from intended goals. The goal is to provide provable guarantees about how AI agents will behave collectively, rather than relying solely on detecting unsafe behavior after it occurs.
“This award gives us the opportunity to shift the paradigm from simply detecting these unsafe behaviors after they occur to providing strict mathematical guarantees that they cannot arise in the first place,” Paleja said.
The research will launch a new direction within Paleja’s Strategies for Collaboration, Autonomy, Learning, and Exploration (SCALE) Robotics Lab, which studies how robots interact and learn with and from other agents. The lab’s research spans the fields of imitation learning, multi-agent reinforcement learning, explainable artificial intelligence, and human-robot interaction.
During the first year of the project, Paleja and his team will develop tools to analyze how AI agents interact and build an understanding of how these interactions can lead to collective failure. This work will ultimately help researchers identify the dynamics and features that contribute to unsafe coordination.
The research could have implications for a variety of complex multi-agent applications, including large-language model collaboration, logistics, cyber defense, and autonomous robotic swarms.
“Complex multi-agent missions remain vulnerable to potentially catastrophic emergent behavior,” Paleja said. “We want to develop the tools to provide meaningful assurances about these systems before those failures can occur.”
The DARPA Young Faculty Award (YFA) program supports early-career researchers at U.S. academic and nonprofit institutions whose work aligns with national security needs. In addition to research funding, the program provides awardees with mentorship and connections to industry and national security organizations. Since its establishment in 2006, the program has supported more than 500 researchers nationwide.
About the Department of Computer Science at Purdue University
Founded in 1962, the Department of Computer Science was created to be an innovative base of knowledge in the emerging field of computing as the first degree-awarding program in the United States. The department continues to advance the computer science industry through research. U.S. News & World Report ranks the department No. 14 and No. 15 overall in undergraduate and graduate computer science, respectively. Graduates of the program are able to solve complex and challenging problems in many fields. Our consistent success in an ever-changing landscape is reflected in the record undergraduate enrollment, increased faculty hiring, innovative research projects, and the creation of new academic programs. Learn more at cs.purdue.edu.