A new paper on arXiv describes a potential attack on LiDAR sensors, which are commonly used in autonomous vehicles and other safety-critical systems. The researchers claim to have identified a previously unexplored attack surface where malware can lie dormant within the LiDAR sensing pipeline, remaining inactive until triggered.
The attack is said to work by sending modulated signals to the LiDAR system, which remotely activate the embedded Trojan malware. The abstract does not provide technical details on the signal characteristics or the activation mechanism, but the authors position this as a new class of threat for autonomous systems.
Because the paper is only available as an abstract, the full methodology and potential countermeasures are not yet described. Still, the finding highlights the need to consider malicious code hidden in sensor processing chains, not just in software or network layers.