Confidential computing aims to protect data while it is being processed (data-in-use). TEEs, which are secure areas within a device's processor, can isolate sensitive computations and data (e.g., cryptographic keys, AI model processing on video feeds) from the rest of the system, protecting them even if the device's main operating system is compromised. Technologies such as Intel's SGX, ARM TrustZone, and AMD SEV provide hardware-backed security guarantees for sensitive operations. In surveillance contexts, TEEs can secure real-time video analytics, ensuring that facial recognition or behavioral analysis happens in a protected enclave inaccessible to potential attackers. This approach addresses privacy concerns by performing sensitive analytics at the edge without exposing raw data. Major cloud providers are now offering confidential computing services that extend these protections to cloud-based surveillance infrastructure. The deployment of TEEs across an IoT surveillance network creates a secure processing chain from capture to storage, significantly raising the bar for would-be attackers. Implementation challenges include performance overhead, complex attestation mechanisms, and the need for careful enclave design to avoid side-channel vulnerabilities.