2009 ushered in mobile malware with the first (and second) iPhone worm appearing just before Christmas. I strongly suspect that 2010 will be a pivotal year in the development of mobile malware. Smartphone software platforms and devices have become sophisticated and flexible enough to allow self-propagating malware and in a world where 9-year-olds are seen holding BlackBerries, the exposure surface is certainly big enough.
Mobile malware still elicits major headlines and is at the center of urgent research as well as speculation. The latest "suspect" came in the form of a banking application that briefly appeared and then was yanked from the Android application market. Security researchers, not having access to the deleted application, are speculating that it was a phishing app that fooled users into installing it as a gateway to banking sites in order to steal credentials and drain bank accounts. Others say it was probably harmless. Unfortunately, since the application was quickly pulled, there is no way to know exactly what it did.
The really interesting story emerges from the fundamental difference in the way Google and Apple deliver applications to their respective devices. While Apple maintains a closed application market with prior vetting of all applications, Google has an open market without vetting. The argument for vetting and signing of applications has always been that a closed application ecosystem made for a more secure device. The problem with that argument is that it gives both the device maker and the carrier ample opportunity to smother any application that threatens their business model (e.g. VoIP). If the vetting process is not transparent, the problem is further exaggerated by seemingly capricious, contradictory or arbitrary decisions by the vetting agent. Already, the Apple marketplace has been the center of multiple application vetting controversies and lawsuits, usually around applications that either offended Apple's sensitivities or threatened its revenue.
In my opinion a-priori vetting of applications is not the best approach to security, especially since it lulls both users and vendors into a false sense of security. Who needs security controls in the operating system all the applications are vetted? The downside of a close application market is that it acts as a drag on innovation without substantial or scalable security benefits.
The good news is that we won' have to wait too long to see how the two divergent strategies work out. The Android and iPhone ecosystems will be repeatedly tested in 2010 and beyond with more malware emerging for both platforms. One application marketplace will carefully vet each app and slow things down, hopefully gaining a security bonus. The other will encourage more innovation and variation but at the potential cost of security. In both cases, malware will slip through and the mobile operating system itself will be tested. Smart users will depend on neither vendor to keep them safe and will opt for operating system controls such as antivirus. And we will get better data for the debate about open vs. closed software ecosystems.