R1 research universities are expanding and reshaping hybrid infrastructure to manage AI-driven volatility in lab computing demand. The report highlights how AI-powered experiments can create unpredictable spikes in IT demand, stressing campus networks and driving pressure on centralized IT teams. At the University of Montana and other R1 peers, on-premises research clusters remain part of the backbone as institutions balance throughput, security, and cost. The focus is on keeping performance stable for both research staff and students while adapting to AI workloads that are less predictable than traditional compute cycles. For campus CIOs, the shift is less about deploying a single AI application and more about building durable capacity planning, monitoring, and governance across campus networks and research clusters as AI changes traffic patterns.