Baptiste Courme adjusts a laser in a Paris lab, where light doesn’t just travel—it remembers. Inside a tangle of mirrors and modulators, his team has achieved something once thought impossible: sending a clear image through a medium that should scramble it completely, using only the ghostly link between entangled photons. At the Institut des NanoSciences de Paris, in collaboration with the Kastler Brossel Laboratory and the University of Glasgow, researchers have turned optical chaos into a filter that opens only for quantum light. This isn’t just about clearer imaging—it’s about redefining how information moves through disorder.
In biological tissues, fog, or twisted optical fibers, light scatters like rain on a windshield, obliterating any image it carries. Traditional methods try to reverse this scattering by carefully shaping light waves using spatial light modulators (SLMs), but they treat all light the same. The breakthrough here is subtler and more powerful: instead of fighting the chaos, the team embraced it, using quantum entanglement as a key. By optimizing a phase mask on an SLM, they preserved the spatial correlations of entangled photon pairs as they passed through a disordered medium—while classical light, sent through the same path, was completely destroyed. The medium, once a barrier, became a gatekeeper, allowing only quantum-encoded information to pass.
The results, published in Nature Physics and Optica in 2026, reveal a system where entanglement isn’t just a curiosity—it’s a functional tool. The team demonstrated that spatial correlations between photon pairs survive the scattering process because entanglement maintains coherence across different optical bases. This means the image encoded in the quantum light emerges intact, while classical light, lacking such correlations, fails entirely. The method relies on a quantum property called “double linearity,” which enables optimization paths impossible in classical systems—paths that resemble solving complex physical problems akin to minimizing energy in multi-spin systems.
The implications ripple across fields. For secure communications, this physical discrimination between classical and quantum signals offers a new layer of protection—eavesdroppers using classical detectors would see nothing. In biomedicine, the technique could lead to non-invasive imaging through tissue without the need for intensive computational reconstruction. And in computing, the optimization process itself may help tackle “hard” problems that stump even the most advanced algorithms.
This isn’t just a lab experiment—it’s a shift in perspective. Complex media are no longer obstacles, but programmable components in a quantum-enabled world. As research continues, the line between noise and signal blurs, not because information is lost, but because we’ve finally learned how to listen in the right way.
