ETH Zurich researchers have introduced a sensor chip that could make manipulated media far easier to detect by attaching a cryptographic signature to images, video, and audio at the instant of capture. The technology, detailed in a paper published in Nature Electronics on March 24, 2026, is aimed at strengthening trust in digital content as deepfakes become more convincing and more widespread.
Rather than securing files only after they have been recorded, the new approach builds verification directly into the sensing hardware. The chip creates a unique digital fingerprint of the captured data and encrypts it using a secure key stored inside the device. That process establishes a traceable connection between the content and the physical sensor that produced it, allowing later checks to confirm when the material was captured, where it came from, and whether it has been modified.
ETH Zurich researchers say this design makes tampering significantly harder. “If data is signed the moment it is captured, any later manipulation leaves traces,” said Fernando Cardes, a research associate at ETH Zurich’s Professorship of Biosystems Engineering and one of the developers of the system. He added that altering the content without detection would likely require a direct physical attack on the chip itself, making large-scale fabrication of false media far less practical.
To support verification, the signatures could be logged by device makers in a public and tamper-resistant ledger, including blockchain-based systems. That would allow third parties to independently confirm whether a file is authentic.
The project began in 2017 at ETH Zurich’s Bio Engineering Laboratory, initially as a side effort tied to the development of highly sensitive sensors for measuring electrical signals in living cells. Researchers involved in that work saw early on that manipulated digital media could become a major problem and started exploring how cryptographic protections could be embedded at the sensor level.
“Trust in digital content is eroding. We wanted to create a technology that gives people a way to verify whether something is genuine,” said Felix Franke, who helped develop the chip at ETH Zurich and now serves as a professor at the University of Basel.
The version described in the paper is a functional prototype intended to prove the concept works in practice. ETH Zurich has filed a patent application and is now looking at ways to make the technology cheaper for camera and sensor manufacturers to adopt. The researchers say the method could, in principle, be integrated into a wide range of devices, and may eventually give platforms a way to automatically check the authenticity of uploaded media.
The work received funding from the Swiss National Science Foundation and the State Secretariat for Education, Research, and Innovation through the SwissChips initiative.
