Yudun treats the on-device model, inference runtime, API credentials, inputs and outputs, and update chain as separate risks. Valuable client code and resources are protected while privileged keys and final authorization remain on the server.
Define client protection for model resources, pre/post-processing logic, and inference invocation paths.
Move long-lived privileged keys off the client and use short-lived, limited credentials with server authorization.
Validate model, runtime, and app version pairing, with disable, staged rollout, and rollback paths.
A public Yudun architecture scenario explains why hiding a string is not a complete control for mobile AI API theft and abuse.
View the risk-control architectureScope: This is a public risk scenario and architecture, not a customer incident or a claimed block rate or cost saving.
Split model, runtime, credentials, interfaces, and user data instead of invoking everything model encryption.
Decide what stays on device and which keys and high-risk actions belong on the server.
Test version pairing, interface permissions, emergency disable, and rollback.
Break down model files, inference runtimes, API credentials, server-side authorization, integrity signals, and logging responsibilities to define a layered security architecture and validation checklist for mobile AI applications.
No. The model must be used at runtime, while inputs, outputs, memory material, interfaces, and credentials remain exposed to separate risks.
Do not treat a long-lived high-privilege key as a client secret. Prefer a server proxy with least privilege, rotation, and request records.
No. On-device inference can reduce latency and improve privacy, but it adds model delivery, version, device resource, and runtime material concerns.
Only when the candidate version, method, covered scenarios, and open items are explicit. Design work or static observation is not runtime verification.
On-device model integration and runtime boundaries
On-device inference runtime and model delivery context
Mobile application security controls and verification scope
Android application security design and release boundaries
Apple platform code signing and runtime security context
御盾