The field of control systems is moving towards a more data-driven approach, with a focus on establishing statistical guarantees for stability and safety. Researchers are exploring innovative methods to quantify uncertainty in model predictions and develop control strategies that can adapt to changing dynamics. One notable direction is the use of conformal prediction and Koopman operator theory to improve the robustness of control systems. Another area of research is the development of online learning control strategies that can learn from real-time data and update their parameters to ensure safe operation. Some papers are also investigating the use of machine learning techniques, such as Gaussian process bandit algorithms and Bayesian optimization, to improve control performance in complex systems. Noteworthy papers include: Statistical Guarantees in Data-Driven Nonlinear Control, which introduces the concept of conformal robustness for stability and safety guarantees. Learning Safe Control via On-the-Fly Bandit Exploration, which presents a method for collecting additional data on the fly to ensure safety in control tasks with high model uncertainty. Multi-Timescale Dynamics Model Bayesian Optimization for Plasma Stabilization in Tokamaks, which demonstrates a comprehensive solution for controlling complex real-world systems using a multi-scale Bayesian optimization approach.