Neuromorphic computing is a brain-inspired approach for processing large amounts of data with high energy efficiency, low latency and high bandwidth. Neuromorphic systems are particularly attractive for edge computing applications thanks to their high performance and lightweight features. However, strong and energy-efficient security layers are mandatory for such systems, especially for safety-critical applications, as they are very easily compromised at the edge. The EU-funded NEUROPULS project aims to build low-power, secure edge computing systems by developing novel photonic computing architectures and security layers based on photonic physical unclonable functions in augmented silicon photonics platforms compatible with CMOS technology. By developing novel technological, hardware and simulation platforms, NEUROPULS aims to create next-generation neuromorphic accelerators featuring RISC-V compliant interfaces for easier system compatibility.