Skip to content

data protection

AI security interview questions tagged data protection, across every topic.

7 questions · 2 unlocked for you

Concepts behind "data protection"

Explanations of the mechanisms these questions rely on.

Foundational
Mobile Security
Protecting Data on Mobile DevicesMobile apps protect data by storing as little as possible, keeping keys in the Android Keystore or iOS Keychain and Secure Enclave, encrypting what must persist with those keys, and choosing protection classes that keep secrets out of backups and inaccessible while the device is locked. Hardware backing stops keys being copied off the device, but not their use by a compromised app. Most real leaks bypass storage entirely: logs, crash reports, analytics SDKs, backups, app-switcher screenshots, the clipboard, keyboard caches, URLs and notifications.
Foundational
Infrastructure Security
Protecting Model WeightsTrained model weights can represent enormous investment and, for the most capable models, real misuse risk if stolen. Theft needs only one successful copy, through compromised credentials, insiders, supply chain attacks, exposed storage or the serving infrastructure itself. Protection combines limiting where weights exist and who can access them, strong identity and encryption with tightly controlled keys, monitored and restricted egress, hardened serving, and insider-risk controls. RAND's 2024 study frames the problem as security levels matched to increasingly capable attackers.
Core
LLM Application SecuritySign in
Sensitive Information DisclosureSensitive information disclosure, LLM02 in the OWASP Top 10 for LLM Applications (2025 and 2026), covers personal data, secrets, proprietary information and other users' data leaking through an LLM application. The leak paths include retrieval without authorization, context shared across users, prompt and response logs, caches, fine-tuning data, model memorisation, third-party providers and the answer itself. Controls minimise what enters each path and enforce access before data reaches the model.