Six ways to prevent a breach like the one at AT&T
- 18 June, 2014 22:42
A data breach like the one recently reported by AT&T demonstrates that security policies alone are only a paper tiger without the technological teeth to make sure they are enforced, experts say.
AT&T reported last week that unauthorized employees of one of its service providers accessed the personal information of AT&T wireless customers. The exposed data included social security numbers and call records.
AT&T did not say how many records were accessed, but the number was high enough that the carrier had to report the breach to California regulators.
While there was no indication of criminal intent, the service provider's employees "violated our strict privacy and security guidelines," AT&T said.
If that meant the carrier was banking only on policy to keep partners in check, then AT&T had made a big mistake.
"Why would you only use policy to limit their access to data?" Steve Hultquist, chief information officer for RedSeal Networks, said. "Wouldn't you have authorization controls as well?"
Experts say there are many ways to place data restrictions on business partners or anyone else accessing a corporate network.
Sometimes the application being accessed can be configured to segment data, so that only people with the proper credentials can view the most sensitive information, Wolfgang Kandek, chief technology officer for risk management company Qualys, said.
Tools that are increasingly being used include "anomaly detection," Kandek said. For example, software is available that can track how many records are being accessed by a service provider per hour and alert when the number crosses a threshold.
Another option is access controls based on the role of the person on the network, Farbod Foomany, consultant and researcher for Security Compass, said. Customer data can be restricted to people identified as having permission to view the information.
"If they (companies) implement a granular access control, that will mitigate the risk of having someone access the data they shouldn't," Foomany said.
Monitoring and logging access to the database is another way of spotting a breach. Software should be in place to log who creates a field, who views it and who modifies it.
"By using that monitoring system, companies can find out who has violated a policy and go and ask why it has happened," Foomany said.
Hultquist believes sensitive data should be segmented, so a service provider doing an unrelated task "can't see stuff they shouldn't be able to see."
Close monitoring of network users requires a company to take on the difficult task of identifying all the Internet connections that could be used as potential access points. Finding all those connections can be "really hard for a lot of companies," Hultquist said.
Once the access points are known, then software can be used to track the users on the network and restricts them to approved paths for accessing data and applications.
Having such technical controls in place to enforce security policies is key to good security, experts say.
"You can't just have some policies on paper," Foomany said. "You have to implement them through security mechanisms."