Last month, an AT&T security loophole allowed the e-mail addresses of 114,000 iPad 3G subscribers, including those of White House Chief of Staff Rahm Emanuel, journalist Diane Sawyer and Dow Jones CEO Les Hinton, to be exposed.
How can such a blatant breach occur? Simple. Private information was stored in the cloud, but the users had no control over how that information was protected. Businesses should not allow any personal information to be stored in the cloud that they don’t have an actionable SLA for. User credentials should be stored within the respective enterprise, with appropriate controls.
This doesn’t preclude businesses from consuming cloud or SaaS services in any way. There are a number of ways cloud providers can deliver services, without putting customers’ data at risk.
Leveraging identity federation
New capabilities are emerging that enable cloud providers to deliver a host of valuable services via real-time mapping and just-in-time account provisioning, while keeping data safely behind its customers’ firewalls. For example, AT&T may have been storing users’ e-mail addresses in order to authenticate them, yet this could easily be done by authenticating via a SAML assertion verifying the user’s role, without revealing their e-mail address. In this scenario, mapping ties the real user to their anonymous identity within AT&T, but the identities are only connected together for the few seconds it takes to authenticate that user. This way, if the information stored by AT&T is ever exposed, there’s no chance of it being tied back to an individual customer.
This involves taking identity federation one step further. Instead of merely using identity federation for tasks like single sign-on, enterprises can share necessary information with cloud providers, without compromising that information.
Shifting the data capture mindset
However, cloud security breaches can’t be solved simply by installing a security product. It requires a change in mindset. Providers are capturing far more customer data than is necessary to perform essential tasks. It’s natural for any cloud provider to desire a wealth of information about its customers to deliver a more personalized experience, yet they usually don’t need all this data in the cloud to accomplish the task at hand, and doing so puts their customers at risk. Using role mapping, they can still identify users’ personal data to customize their experiences, without exposing this data in the cloud.
For example, each iPad 3G has an AT&T SIM card within it that stores a certain amount of information, so why store that same data in an AT&T database, when they can get the data in real time, just by pinging the user’s iPad? AT&T may want to store my information in a separate database so they can send me customized emails, for example, but as a customer I should be able to opt out of having my personal information stored in this way. Otherwise I have no idea how you’re protecting my e-mail address and other sensitive data.
What are your concerns about cloud security? And how are your cloud providers protecting your user identities today?
Posted
Aug 05 2010, 09:57 AM
by
BaberAmin