Physical tampering is direct manipulation of hardware to change how it behaves. Instead of attacking software over a network, an attacker touches the device itself: opening the case, probing circuits, altering voltage or timing, injecting light or electromagnetic effects, or otherwise disturbing the chip while it runs. In embedded security, this matters because many protections assume the device will operate under normal conditions.
Tampering can target debug ports, secure boot checks, key storage, or fault-sensitive logic. If successful, it may expose firmware, bypass safeguards, or push a chip into an unexpected state. Defenders respond with tamper-resistant packaging, locked debug interfaces, secure boot, TrustZone or similar isolation, sensor-based response, and careful threat modeling for physical access. The key idea is that a device is not fully secure if an attacker can still influence its hardware behavior at close range.