Unpack mobile app security risks with these in-depth guides | Promon

When AI attacks the app (and what you can do about it)

Written by Promon | Nov 20, 2026, 12:13:59 PM

What changes when AI attacks your app?

AI has changed the economics of app attacks. Models now find decades-old vulnerabilities that human reviewers missed, AI agents drive reverse-engineering tools such as Frida directly, and banking trojans such as RemControl and PromptSpy use generative AI to build infrastructure and automate actions on infected phones. At the same time, teams are shipping AI models, prompts, and decision logic inside their apps, on devices attackers control.

But AI hasn't changed where attacks happen. Whether a human or an AI runs them, attacks still rely on techniques inside the device: abused Accessibility services, injected input, hooked code, deepfakes, and app tampering.

This whitepaper explains what AI-assisted attacks look like in the field today, why defense must live inside the app, and how the Promon Runtime Intelligence Loop learns from attacks globally and adapts protection locally. It closes with five practical steps to reduce your exposure.

What you'll learn:

How AI is already used in real mobile attacks, from RemControl and PromptSpy to deepfakes injected into banking KYC flows.

Why on-device AI models and shipped IP are now targets, and what 49 threats to on-device AI look like.

How well ten leading AI models deobfuscate protected code, and where layered obfuscation holds.

Why signature scanning falls short, and how to detect the technique instead of the file.

Five steps your team can take now to protect apps, APIs, and AI at runtime.


Are your apps ready for AI-assisted attacks?

AI is changing how attackers find vulnerabilities and execute attacks, but the need to protect the running app remains.

Download the whitepaper to understand today's AI-assisted threats and discover five practical steps to reduce your exposure.