Fiber Optic Networks Can Wiretap Nearby Speech, Seismic Research Demonstrates

Researchers have demonstrated that fiber optic cables—widely deployed for telecommunications and increasingly used for earthquake monitoring—can be weaponized as covert acoustic surveillance devices. The technique, known as distributed acoustic sensing (DAS), exploits how laser pulses traveling through fiber optic strands reflect off microscopic imperfections. When vibrations from nearby speech reach the cable, they alter these reflections in patterns that, with sufficient processing, yield intelligible audio.

In controlled field tests, scientists placed speakers adjacent to standard fiber optic cables and played tones, music, and human speech. By analyzing the reflected light patterns and feeding the data through publicly available AI transcription software, the team successfully reconstructed real-time speech transcripts from the cable vibrations alone. The findings suggest that existing seismic monitoring infrastructure—deployed across earthquake-prone regions and urban centers—carries an underappreciated dual-use capability.

The eavesdropping range, however, faces significant physical constraints. Speech recovery proved effective only with exposed, coiled cable configurations within approximately five meters of the sound source. Burying the cable beneath just 20 centimeters of soil substantially degraded audio quality, and straight cable runs demonstrated poor performance. These limitations narrow the practical threat surface but raise pointed questions about privacy exposure wherever fiber optic networks run near exposed or accessible infrastructure. The research highlights how readily available sensing technology, combined with advancing AI transcription tools, could lower the barrier to covert audio surveillance in specific deployment scenarios.