We report recent progress on the use of generative artificial intelligence (AI) methods to accelerate particle-based plasma kinetic computations. These computations are time-consuming due to multiscale dynamics, boundary conditions, and the need to follow large ensembles of particles to avoid statistical sampling errors. The physics models of interest are Fokker-Planck (FP) equations for the particle distribution function in phase space including drifts, diffusion, and collisions. The AI methods include Normalizing Flows (NF) and Diffusion Models (DM). We present a pseudo-reversible NF model that learns the distribution of the final state conditioned to the initial state, such that the model only needs to be trained once and then used to handle arbitrary initial conditions [1]. Following this, we present results based on the use of DM that allow the quantification of confinement losses in bounded domains. We propose a unified hybrid data-driven approach that combines a conditional DM with an exit prediction neural network to capture both interior stochastic dynamics and boundary exit phenomena [2]. Convergence analysis, along with numerical test experiments are provided to demonstrate the effectiveness of the proposed methods. We present applications to advection-diffusion transport in 3D chaotic flows, and the generation and confinement of runaway electrons in magnetically confined fusion plasmas. [1] M. Yang et al, SIAM Journal of Scientific Computing, 46, (4) C508-C533 (2024). [2] M. Yang et al, Submitted to J. Comp. Phys. arXiv:2507.15990v1 (2025).