Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Mobile authentication fails in predictable ways: SMS codes intercepted by SIM swap, tokens stored in the clear, biometrics checked client-side, and OAuth flows run inside a WebView.

What Are Mobile Authentication Risks?
Authentication on a phone has to work in three seconds, in one hand, on a device the user may have jailbroken and an attacker may be holding. Those constraints push developers towards shortcuts, and the shortcuts are where assessments keep finding the same handful of weaknesses.
SMS one-time codes remain the most widely deployed second factor and the weakest in common use. SIM swap fraud—persuading or bribing a network to port a number to a new SIM—redirects every code to the attacker, and UK Finance has repeatedly flagged impersonation fraud of this kind in its annual reporting. On the device itself, malware with the SMS read permission or an abused accessibility service captures codes without touching the network at all.
Push-based approval is better but not immune: prompt bombing, where an attacker repeatedly triggers approvals until a tired user accepts, worked well enough to feature in several high-profile 2022 intrusions. Number matching, where the user types a digit shown on the sign-in screen, closes most of that gap. Phishing-resistant options—passkeys and platform authenticators under WebAuthn—remove the shared secret entirely and are now realistic on both platforms. See multi-factor authentication for how the factor types compare.
Once a user signs in, the app holds something valuable. Testers routinely find refresh tokens in SharedPreferences, session identifiers in a plist, JWTs written to logs, and credentials cached so the user never has to type them again. Anything outside the Android Keystore or the iOS Keychain is readable on a rooted device, in a backup, or occasionally from an unencrypted log buffer.
The supporting mistakes are equally common: refresh tokens that never expire, no revocation path when a device is reported lost, and no binding between the token and the device it was issued to. Hardware-backed key storage with authentication binding fixes most of this—the key is unusable until the user passes a biometric or PIN check.
A biometric prompt is not authentication in itself; it is a local gate. If the app simply calls the API and branches on a boolean, a hooking framework such as Frida flips the branch in seconds. The correct pattern binds the biometric to a cryptographic operation: on Android, a BiometricPrompt.CryptoObject wrapping a Keystore key created with setUserAuthenticationRequired(true); on iOS, a Keychain item with an access control flag requiring biometry. The decryption only succeeds after a genuine match, so there is no boolean to patch.
Also check the authenticator class. Android distinguishes BIOMETRIC_STRONG from BIOMETRIC_WEAK, and face unlock on many handsets is weak—fine for convenience, not for authorising a payment.
RFC 8252 is unambiguous: native apps should run OAuth in the system browser—Custom Tabs on Android, ASWebAuthenticationSession on iOS—not in an embedded WebView, where the app can read the credentials the user types and the user cannot see the real address bar. PKCE is mandatory for public clients, because a mobile app cannot keep a client secret, and the redirect URI must be an app link or universal link with domain verification rather than a custom scheme any other app can claim. Weak redirect handling turns an authorisation code into something another installed app can steal.
Overlay attacks draw a fake login screen on top of the real app, and accessibility-service malware reads the fields directly. Android banking trojans such as Anatsa and Hook have built entire business models on this, as covered in our article on mobile malware. Setting FLAG_SECURE on sensitive screens, detecting overlay conditions, and refusing to run alongside an unexpected accessibility service all raise the cost, and server-side attestation through the Play Integrity API or Apple’s App Attest gives you a signal that is not forgeable on the client.
For payments, strong customer authentication under the Payment Services Regulations 2017 requires two independent elements plus dynamic linking of each transaction to its amount and payee—so a generic code that merely proves a login is not sufficient. The NCSC’s guidance continues to recommend moving away from SMS where a stronger factor is available, and FCA-regulated firms are expected to show the reasoning behind whatever they deploy.
At Privia Security, our Mobile Application Security Services test these flows end to end, from the biometric prompt to the token your API accepts.
It is better than a password alone and far weaker than the alternatives. SIM swap, network interception, and on-device malware all defeat it, and it offers no protection against real-time phishing proxies. Where the account protects money or sensitive personal data, move to an authenticator app with number matching, or to passkeys. Keep SMS as a fallback for users who have no other option.
In platform-provided secure storage: the Android Keystore or the iOS Keychain, with an accessibility class that keeps the item on the device and requires an unlocked state. Bind the key to user authentication where the risk justifies it, keep refresh token lifetimes short, and provide a server-side revocation path so a lost device can be cut off without waiting for expiry.
The sensor is rarely the weak point. The common bypass targets the app logic: if the code only checks whether the biometric API returned success, that result can be hooked on a rooted device. Binding the biometric to a cryptographic operation on a hardware-backed key removes the shortcut, because the data cannot be decrypted without a genuine match.
An embedded WebView is controlled by the host app, so it can read the credentials typed into it, and the user cannot verify the address or see the certificate. RFC 8252 therefore requires native apps to use the system browser through Custom Tabs or ASWebAuthenticationSession, which keeps the session isolated from the app and lets the user see exactly which site they are signing in to.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.