Indistinguishability in probability refers to the idea that two stochastic processes are indistinguishable if they are almost surely equal at every point in time, meaning the probability that they differ at any point is zero. This concept is crucial in stochastic analysis, especially in the study of martingales and Brownian motion, where it ensures that different sample paths of a process are effectively the same in terms of probability theory.