Mobile Top 10 2012 - M3 Insufficient Transport Layer Protection

When designing a mobile application, commonly data is exchanged in a client-server fashion. When this data is exchanged it traverses both the carrier network and the internet. For sensitive data, if the application is coded poorly, threat agents can use techniques to view this sensitive data while it's travelling across the wire. The following threat agents exist:

 The exploitabilty factor of monitoring a network for insecure communications ranges. Monitoring traffic over a carriers network is harder than that of monitoring a local coffee shops traffic. In general targeted attacks are easier to perform.  Unfortunately, mobile applications frequently do not protect network traffic. They may use SSL/TLS during authentication, but not elsewhere, exposing data and session IDs to interception. In addition, the existence of transport security does not mean it is implemented to it's full potential.
 * Users local to your network (compromised or monitored wifi)
 * Carrier or network devices (routers, cell towers, proxys, etc)
 * Malware pre-exisiting on your phone

Detecting basic flaws is easy. Just observe the phone's network traffic. More subtle flaws require inspecting the design of the application and the applications configuration. Such flaws expose individual users’ data and can lead to account theft. If an admin account was compromised, the entire site could be exposed. Poor SSL setup can also facilitate phishing and MITM attacks. Consider the business value of the data exposed on the communications channel in terms of its confidentiality and integrity needs, and the need to authenticate both participants.

The best way to find out if an application has sufficient transport layer protection is to look at the application traffic through a proxy. Find the answers to the following questions:


 * 1) Are all connections, not just ones to servers you own, properly encrypted?
 * 2) Are the SSL certificates in date?
 * 3) Are the SSL certificates self signed?
 * 4) Does the SSL use high enough cipher strengths?
 * 5) Will your application accept user accepted certificates as authorities?

General Best Practices:


 * Assume that the network layer is not secure and may potentially be hostile and eavesdropping.
 * Enforce the use of SSL/TLS for all transport channels in which sensitive information, session tokens, or other sensitive data is going to be communicated to a backend API or web service.
 * Remember to account for outside entities like 3rd party analytics companies, social networks, etc, and use their SSL versions even when an application runs a routine via the browser/webkit. Mixed SSL sessions should be avoided and may expose the user’s session ID.
 * Use strong, industry standard encryption algorithms and appropriate key lengths.
 * Use certificates signed by a trusted CA provider.
 * Never allow self-signed certificates, and consider certificate pinning for security conscious applications.
 * Do not disable or ignore SSL chain verification.
 * Only establish a secure connection after verifying the identity of the endpoint server with trusted certificates in the key chain.
 * Alert users through the UI if an invalid certificate is detected.
 * Do not send sensitive data over alternate channels, such as SMS, MMS, or notifications.

iOS Specific Best Practices

Default classes in iOS handle SSL cipher strength and negotiation very well as of recent releases. The trouble comes when code is introduced to bypass these defaults to accommodate development hurdles. In addition to the above general practices:


 * Ensure that certificates are valid and fail closed.
 * When using CFNetwork, consider using the Secure Transport API to designate trusted client certificates. In almost all situations, NSStreamSocketSecurityLevelSSLv3 or NSStreamSocketSecurityLevelTLSv1 should be used for higher standard cipher strength.
 * After development ensure all NSURL calls (or wrappers of NSURL) do not allow self signed or invalid certificates such as the NSURL class method setAllowsAnyHTTPSCertificate.
 * Consider using certificate pinning by exporting your certificate, including it in your app bundle, and anchoring it for your trust object. Using the NSURL method connection:willSendRequestForAuthenticationChallenge: will now accept your cert.

Android Specific Best Practices


 * Remove all code after the development cycle that may allow the application to accept all certificates such as org.apache.http.conn.ssl.AllowAllHostnameVerifier or SSLSocketFactory.ALLOW_ALL_HOSTNAME_VERIFIER. This is equivalent to trusting all certificates.

Example Scenarios


 * Fortify On Demand Blog - Exploring The OWASP Mobile Top 10: M3 Insufficient Transport Layer Protection
 * OWASPIOS Developer Cheat Sheet
 * blog.securemacprogramming.com - SSL Pinning for Cocoa Touch
 * Why Eve and Mallory Love Android: An Analysis of Android SSL (In)Security
 * Fortify VulnCat - A Taxonimy of Software Security Errors