Memory loss is contagious in open quantum systems

Markovian (memoryless) system-bath interactions play a fundamental role across physics, chemistry, and biological systems. Typically, such memoryless behavior arises from the bath’s properties. Here, we reveal a distinct mechanism: when a system interacts with both a Markovian and a non-Markovian bath, losses induced by the former suppress memory effects from the latter, making the interaction appear more Markovian. This effect leads to a direct interplay between independent baths, where Markovianity becomes “contagious,” transferring between them via their common system. To describe this phenomenon, we introduce a Bloch-Redfield-inspired approach that accurately captures how a lossy system, governed by a Lindblad master equation, interacts with a non-Markovian bath. Beyond offering new insights into dissipative dynamics, this framework provides a computationally efficient alternative for modeling complex system-bath interactions across a broad range of scientific disciplines. Markovianity, where system-bath interactions have no memory, is central for describing many physical processes and serves as a useful approximation that reduces complexity. Here, the authors show that Markovianity can arise not only from the bath’s properties, but also from dissipation induced in a system coupled to a non-Markovian bath by its coupling to an additional Markovian bath.