Author:

Abstract:

The security of Internet-of-Things (IoT) is a growing concern, with IP cameras like those from Eufy promising robust security through military-grade encryption. While Eufy's claims are strong, independent verification of these claims is crucial to confirm the integrity and resilience of its systems against potential vulnerabilities and extend the lessons learned to the broader IoT landscape, ensuring practices keep pace with technological advancements. We unveiled the inner workings and security measures in the Eufy ecosystem through reverse engineering, particularly focusing on its smart doorbell and Homebase, and evaluated the proprietary peer-to-peer protocol and encryption methods. This paper offers a comprehensive analysis of the Eufy ecosystem, offering insights into the broader implications of IoT device security. Our investigation revealed critical vulnerabilities within the ecosystem, which were responsibly disclosed and confirmed by Eufy. The vulnerabilities could compromise end-user privacy by allowing unauthorized access to the end users' private network within seconds. A key tool in our research was \texttt{dAngr}, a symbolic debugger we developed to facilitate the reconstruction of encryption keys in intricate cross-architecture binaries, thus enabling a more efficient reverse engineering process. The research revealed vulnerabilities in Eufy's ecosystem, leading to serious privacy and security concerns, and suggests effective countermeasures, stressing the need for continued vigilance in IoT device security.