3 Steps To Keep Down Security’s False-Positive Workload
Security needs to be better automated, but while detecting attackers is great, all too often automation means that security teams are left with chasing down a list of security events that turn out not to be an attack but unexpected system, network, or user behavior.
These “false positives” are the bane of most machine-learning systems: valid e-mail messages blocked by anti-spam systems, unexploitable software defects flagged by software analysis systems, and normal application traffic identified as potentially malicious by an intrusion detection system. First-generation security information and event management (SIEM) systems, for example, would often deliver lists of potential “offenses” to security teams, leading to a lot of work in wild goose chases, says Jay Bretzmann, market segment manager for security intelligence at IBM Security Systems.
“If you cannot manage the list of offenses that come into the product in a day, then you need to do some tuning, or you need to go out and do some proactive defense, such as eliminating vulnerabilities by patching or looking at your configurations,” Bretzmann says.
Beating the false-positives problem is key to making the company more secure, and the first step is to not think of such alerts as false positives, says Paul Stamp, director of product marketing at RSA. There is always a reason why a security system flags some behavior as threatening, he says.
“There is no such thing as a false positive,” he says. “It is just that some positives are more important than others.”
To raise the bar on security events, experts recommend a few tactics.
1. Better tuning
The initial training of a security system — whether a network anomaly detector or the log analysis component of a SIEM — is a necessary step toward teaching the appliance what should be considered bad and what’s good on the network.
However, the training is a crash course for the device on what is typically normal, or not, in the network for a short period of time. The training set often includes previously detected malicious behavior at other companies, which are quickly turned into rules and exported to the rest of the client base. Most of the time that helps detect true threats, but sometimes a vulnerability scanner, uncommon user behavior, or other event can set off the security system.
“We can help you identify what is malicious traffic based on what our other customer thought was malicious traffic, but it’s only a start,” IBM’s Bretzmann says. “You really have to get in there and investigate what’s causing the false positives.”
Security administrators should work from a list of the most common types of alerts, or “offenses,” as IBM calls them. Alerts that occur frequently are likely false positives but should be remediated to reduce the noise.
[Companies analyzing the voluminous data produced by information systems should make sure to check user access and configuration changes, among other log events. See 5 Signs Of Trouble In Your Network.]
2. Proactive defense
If tuning fails to remove enough alerts to allow the security team to focus on the most severe events, then the IT security managers should consider proactive work that can reduce vulnerability and allow the rules to be tightened. For example, legacy systems that are not patched regularly can dramatically increase the vulnerable attack surface area of the company, resulting in a higher number of alerts.
“The rate of new vulnerability disclosures ranges between 12 and 15 a day,” Bretzmann says. “Trying to keep up with that is a never ending job.”
Patching vulnerable systems, shutting down unnecessary systems, and limiting the types of network traffic that can enter the network are all steps that can reduce the attack surface area of the business and allow the security systems to be tuned more tightly, reducing the number of alerts.
3. Add more context
Combining external threat intelligence can help focus security administrators on the most important threats, so they can prioritize security events, says Erik Giesa, senior vice president of business development at operational-intelligence provider ExtraHop.
“You want the ability to be very surgical and precise,” he says. “You don’t want it to be a firehose.”
False positives can also be eliminated by knowing more about the assets in the network, such as which systems are most important and which can effectively be ignored for the moment.
IT security should cooperate with business executives to collect data on what information-technology components are core to the business and should be closely watched, RSA’s Stamp says.
“A lot of the knowledge about the risk associated with a system isn’t held by IT — IT doesn’t know, the business knows,” he says. “So the question is how do you involve the business in the process?”
Have a comment on this story? Please click “Add Your Comment” below. If you’d like to contact Dark Reading’s editors directly, send us a message.