Deploying intelligent systems at the edge presents a unprecedented set of risks. These decentralized platforms operate with constrained resources and connectivity, making them vulnerable to a range of threats. From adversarial access to data violations, safeguarding edge AI deployments requires a