Watson for i: Cybersecurity the Watson Way

Analytics & Cognitive - Other
  • Smaller Small Medium Big Bigger
  • Default Helvetica Segoe Georgia Times

The escalating hacks of businesses and the mushrooming of ransomware show that cybersecurity is going to be a bigger part of future IT life than ever before. IBM Watson offers services that can help.

More than a year and a half ago, IBM Chairman Ginni Rometty was quoted in a Forbes report as saying that "cybercrime is the greatest threat to every company in the world." Since then, we've seen hacking causing controversy in the U.S. election for president in 2016, the WannaCry ransomware attacks last May affecting more than 230,000 computers in more than 150 countries, and intrusions at U.S. power plants (including a nuclear facility) in early July resulting in yet-undefined consequences, to name just a few of the most notorious examples. Cyber warfare is here to stay, and no one can be complacent that their systems will remain immune forever.

"A lot of companies have very old software that no one wants to touch, and the companies have decided to let it be and leave it unpatched and unprotected," noted Morningstar analyst Ilya Kundozerov in a recent CNBC article on cybersecurity stocks. Does that situation sound familiar to anyone? You might hope not, but we all know better.

The problem is what to do about it. There are so many potential ways for intruders to sneak into corporate systems that an army of little Dutch boys aren't going to be enough to plug every hole. The security features of the IBM i are helpful, but no solution is perfect, and many companies use additional platforms that are more vulnerable.

IBM estimates that each year there are more than 75,000 software vulnerabilities discovered and more than 10,000 research papers on security challenges published, as well as more than 60,000 blogs on security topics published monthly. No human has time to read all that material, much less do something meaningful with it.

One solution you could use, ideally, is some sort of artificial intelligence (AI) that could draw and summarize information on cyber attacks from these thousands of sources and help IT staffs learn about attacks and potential remedies. Fortunately, such a solution exists.

Keeping Up with the BushwHackers

Watson for Cyber Security is IBM's umbrella term for a group of cybersecurity services that can help IT staffs parse what's going on with their own systems and networks and what's coming out of the security information fire hose. IBM Watson's AI, once trained in security concerns, can process and interpret the latest security data, including that portion of it that's unstructured (i.e., in written and spoken language), using reasoning that mimics human thought. Then actual thinking humans can proactively implement the information that's most pertinent to the security needs of their own organizations. This should provide better results than relying on luck that in-house analysts will track down on their own the most germane security information relating to a particular threat. In particular, it can help analysts differentiate between normal system behaviors and variations that might mean an attack is underway.

QRadar Advisor with Watson is part of IBM's QRadar Security Analytics Platform, which consists of several parts that provide functions other than surveying outside security information. QRadar Advisor draws together the information and expertise provided by the other parts of the Security Analytics Platform to provide a comprehensive picture of potential security threats, their backgrounds, and available remedies. To get an idea of how this happens, it's useful to look at the Security Analytics Platform's various parts.

Components of the QRadar Security Analytics Platform

IBM QRadar User Behavior Analytics (UBA) learns to recognize the characteristics of regular system users and with that knowledge can detect anomalous behavior undertaken by the accounts of those users. UBA summarizes these behaviors with other unusual incidents detected by other QRadar features in a dashboard format that helps spot, for example, authorized users who are violating security policies (e.g., by downloading information they shouldn't) and outsiders who may be trying to commit security offenses while masquerading as authorized users. This information lets analysts, for example, identify high-risk users, log violations of security policies, and use actual incidents as the basis for revising existing security rules.

IBM QRadar Incident Forensics goes a step further and lets analysts review each action taken by a suspicious account, compare it to past actions of that and other users to assess threat levels, and correlate it with other data to quickly determine how much of a problem a particular sequence of events may pose to system security. Incident Forensics optionally works with network capture packet devices (PCAPs) that preserve network traffic records in files for further analysis.

IBM QRadar Network Insights helps analysts make better sense of the network security logs that are generated by any network activity. Normally, these logs are voluminous, which can make finding and interpreting any particular action difficult for a human. Network Insights contains algorithms that can help analysts quickly track down network threats (e.g., malware, phishing) hidden in the data so potential problems can be dealt with before they escalate.

IBM QRadar Network Security (XGS) focuses on monitoring and controlling network activities. It uses techniques such as deep-packet inspection (examination of the data part of each packet to find violations of specific criteria and to collect statistical information), heuristics (logical shortcuts to solve problems that are resistant to normal analysis), and analysis of network and user behavior to detect intrusions or other problems. XGS services, for example, let administrators view SSL/TLS-encrypted packets moving through a network and control user access to specific URLs and applications.

XGS relies on data from the IBM X-Force Exchange, an IBM-sponsored security forum that draws on the experience of numerous experts and data sources in the security field to keep its members abreast of the latest developments and security threats as they appear. Any interested person can join this exchange to keep up with the constantly morphing security threat situation worldwide.

IBM QRadar Vulnerability Manager looks for more than 70,000 known security problems (e.g., default settings, configuration errors, software lapses) and uses a Payment Card Industry (PCI)-certified scanner to detect them while accepting input from other designated outside data sources. Vulnerability Manager uses this data to prioritize security threats, zero in on activities that may herald such threats, and help automate compliance with federal regulations and industry standards. Its automated policy engine can calculate risks and forecast ways in which a particular risk could escalate to threaten an enterprise's information assets.

Watson and the Cognitive SOC

All these components can individually provide useful help for combatting security problems, but overall they contribute to the IBM Cognitive Security Operations Center (SOC), which, as IBM puts it, is "powered by Watson for Cyber Security" and defines QRadar Advisor with Watson as "its centerpiece." What this means is that the Cognitive SOC is a unifying option that brings all the Watson for Cyber Security threads together into a centralized command-center format that can handle security protection for even the largest of enterprises.

This service relies on the concept of security information and event management (SIEM), which is generally defined as combining security information management (SIM) with security event management (SEM). Simply speaking, SIM collects security data into a central data store for automated reporting while SEM controls collection of activity logs and facilitates their rapid interpretation. Taken together, SIEM systems help enterprises determine and prove that they are meeting compliance requirements of laws and industry standards for particular types of businesses, such as credit-card processors and healthcare providers. This consolidation of effort is part of what makes up the Cognitive SOC.

Another piece of the Cognitive SOC picture is IBM BigFix Detect. This service deals with protecting endpoints, which are the connections to applications and networks used by client devices and are a major traditional attack point for intruders. BigFix Detect combines threat detection with compliance policy reporting, software patching and distribution, inventorying of authorized and unauthorized software present on enterprise systems, and automated patching of detected problems. Like the X-Force Exchange, BigFix also has a user community that shares information and experiences.

Watson Cybersecurity for SMBs

Taking all of these services into account makes the Watson for Cyber Security family sound as if it's just for large organizations. It was clearly designed with big organizations with potentially high compliance thresholds to meet in mind, but IBM offers suggestions for downsized packages that can help SMBs. Essentially, IBM recommends finding a local partner that potentially might use some of the Watson services to act as a managed security services provider (MSSP). On the other hand, the Watson offerings are all cloud services, and a smaller organization could pick and choose which of the services fits their perceived threat levels and sources, with or without the help of an MSSP.

You can learn more about all the Watson for Cyber Security options using the hot links above to access additional videos, white papers, and other materials.