The Agentic AI in these apps is implemented quite uniquely, ensuring your data and details are safe. Always.
In our backend the AI is locked in it's own sandbox, with absolutely no access to anything, except for the data that's being fed into it's sandbox. So no matter what prompts users try to use to confuse or jailbreak it. Even if they managed, the AI wouldn't be able anyway. It has no access to keys, system resources, etc.