WWDC26: Secure your app: mitigate risks to agentic features | Apple
Apple details security techniques and APIs to mitigate prompt-injection risks in LLM-powered agentic app features.
“LLM's introduce a new probabilistic engine within your application that is both powerful but risks being tricked.”
At WWDC26, Apple presented guidance on securing agentic app features built with its Foundation Models and App Intents frameworks, focusing on external attackers exploiting LLM-driven actions rather than model safety. It matters because it signals platform-level treatment of prompt-injection and agentic exploitation as a first-class app security concern as LLM features become mainstream on Apple's platforms.