Can quantum circuits help spot fraudulent transactions?

Credit card fraud datasets are lopsided — genuine transactions vastly outnumber fraudulent ones, causing most classifiers to ignore the rare cases that matter most. Q-SYNTH replaces the generator half of a standard GAN with a parameterized quantum circuit to synthesize realistic minority-class examples, then measures how well those fakes improve downstream fraud detection. The quantum approach closes the gap with classical GANs on distributional fidelity but doesn't yet beat them outright, making it a proof-of-concept rather than a drop-in replacement — though the tradeoff looks promising as quantum hardware matures.