Enterprise -> Technology
By: Dana Gardner, Principal Analyst, Interarbor Solutions
Published: 5th October 2012
Copyright Interarbor Solutions © 2012
The next edition of the HP Discover Performance podcast series unpacks the concept of intelligent containment of risk as an important approach to overall IT security. We'll examine why rapid and proactive containment of problems and breaches—in addition to just trying to keep the bad guys out of your systems—makes sense in today's environment.
Here to share his perceptions on some new ways to better manage security from the vantage of containment is our guest, Kaivan Rahbari, Senior Vice President of Risk Management at FIS Global, based in Jacksonville, Florida. The discussion is moderated by Dana Gardner, Principal Analyst at Interarbor Solutions. [Disclosure: HP is a sponsor of BriefingsDirect podcasts.]
Here are some excerpts:
Gardner: What's different about the overall security landscape today, compared to five years ago?
Rahbari: A lot has changed in the past five years. Two key economic trends have really accelerated our security changes. First, the US recession pushed companies to consolidate and integrate technology footprints and leverage systems. New deployment models, such as software as a service (SaaS) and cloud, help address some of the lack of capital that we've been experiencing and the ability to push cost from fixed to variable.
The second major economic trend that has continued in the past five years is globalization for some of the companies. That means a network topology that's traversing multiple countries and with different laws that we have to deal with.
We always talk about how we're only as strong as our weakest link. When larger and more sophisticated companies acquire smaller ones, which is pretty commonplace now in the market, and they try to quickly integrate to cut cost and improve service, they're usually introducing weaker links in the security chain.
Strong acquirers now are requiring an acquisition to go through an assessment, such as an ISO 27001 certification, before they're allowed to join “that trusted network." So a lot of changes, significant changes, in the past five years.
Gardner: Tell us about FIS Global.
Rahbari: FIS is a Fortune 500 company, a global company with customers in over 100 countries and 33,000 employees. FIS has had a history in the past 10 years of acquiring three to five companies a year. So it has experienced very rapid growth and expansion globally. Security is one of the key focuses in the company, because we're the world's largest wholesaler of IT solutions to banks.
Transaction and core processing is an expertise of ours, and our financial institutions obviously expect their data to be safe and secure within our environments. I'm a Senior Vice President in the Risk Management Group. My current role is oversight over security and risk functions that are being deployed across North America.
Gardner: What is the nature of security threats nowadays?
Rahbari: Attackers are definitely getting smarter and finding new ways to circumvent any security measure. Five years ago, a vast majority of these threats were just hackers and primarily focused on creating a nuisance, or there were criminals with limited technology skills and resources.
Cyber attacks now are big business, at times involving organized crime. These are intruders with PhDs. There could be espionage involved, and originate in countries with no extradition agreements with the US, making it very difficult for us to prosecute people even after we identify them.
You've also read some of the headlines in the past six months, things such as Sony estimating a data breach and cleanup of $171 million, or an RSA hack costing EMC $66 million. So this is truly a big business with significant impacted companies.
Another key trend during the past five years that we've seen in this area is that the nature of the threats are changing from a very broad, scattered approach to highly focused and targeted. You're now hearing things such as designer malware or stealth bots, things that just didn't exist five or 10 years ago.
Other key trends that you're seeing is that mobility and mobile computing have really taken off, and we now have to protect people and equipment that could be in very hostile environments. When they're open, there's no security.
The third key area is cloud computing, when the data is no longer on your premises and you need to now rely on combined security of your company, as well as vendors and partners.
The last major thing that's impacting us is regulatory environment and compliance. Today, a common part of any security expert terminology are words such as payment card industry (PCI), Gramm-Leach-Bliley Act (GLBA), and Sarbanes-Oxley (SOX), which were not part of our common vocabulary many years ago.
Gardner: So how do we play better defense? Most of the security from five years ago was all about building a better wall around your organization, preventing any entry. You seem to have a concept that accepts the fact that breaches are inevitable, but that focuses on containment of issues, when a breach occurs. Perhaps you could paint a picture here about this concept of containment.
Rahbari: As you said, it's easier to secure the perimeter—just don't let anything in or out. Of course, that's really not realistic. For a vast majority of the companies, we need to be able to allow legitimate traffic to move in and out of our environments and try to determine what should be blocked.
I'll say that companies with reasonable security still focus on a solid perimeter defense. But companies with great security, not only guard their perimeter well, they assume that it can be breached at any time, as you stated.
Some examples of reasonable security would include intrusion protection, proxies to monitor traffic, and firewalls on the perimeter. You would then do penetration testing. On their PCs you see antivirus, encryption, and tools for asset and patch management. You also see antivirus and patch management on the servers and the databases. These are pretty common tools, defensive tools.
But companies that are evolving and are more advanced in that area have deployed solutions such a comprehensive logging solutions for DNS, DHCP, VPN, and Windows Security events. They have very complex security and password requirements.
As you know, password-cracking software is pretty common on the Internet nowadays. They also make sure that their systems are fully patched all the time. Proactively, as you know, Microsoft publishes patches every month. So it's no longer sufficient to upgrade a system or patch it once every few years. It's a monthly, sometimes daily, event.
Gardner: How do you go about such containing?
Rahbari: First, as the costs of attacks have skyrocketed, we're now seeing in the market some pretty great solutions that actively try to prevent things from happening or mitigating them when they happen.
A few of the examples that come to mind are on the perimeter. We're seeing a lot of denial-of-service (DoS) attacks in recent years. Basically what that means is you detect a massive attack toward a specific IP address, and it happens with financial institutions a lot.
With some of these great solutions on the market, you would swing all of your traffic to another IP address, without bringing down the environment. The attackers still think they're attacking and shutting down an environment, but they really aren't.
Five years ago, the primary objective of a DoS attack like this was just to shut something down for malicious purposes. Now, it's a pretty common vehicle for fraud.
Here's a scenario. In a small business, Joe's Landscaping, their internet banking gets compromised and someone steals their password. Then the hackers authenticate and do a wire transfer out of his account to some bank in the Cayman Islands. That attacker then mounts the DoS attack against the service provider that Joe's Landscaping is using, so that the fraudulent activity is not discovered or it's delayed for a few days. By the time it's discovered, the money is long gone.
Another example of proactive and great security is to have software white-listing on PCs and servers, so that only legitimate software is actually installed. A key method of obtaining credentials nowadays is to install keystroke logger software on a machine. That can easily be blocked by white-listing software.
A third example of strong security posture is not just to detect, but actually actively destroy things. Traditionally, when we were monitoring wireless access into companies, we would just report that there was wireless access that shouldn't have been granted.
Most companies would assign a password to it, but as you know, passwords are shared sometimes, so soon enough, everyone knows the password to the wireless system in the company. One of the things that we’ve started using are solutions that actively jam wireless signals, unless it's their authorized room or a known IP address.
Another great example of a proactive approach we see in the market is when a visitor or employee plugs a non-corporate device into network, either on premise or from home. That creates a significant amount of risk. There are some great solutions out there that provide network access control. If an unknown device plugs into the network, that's immediately rejected at the network level. You can't even authenticate.
Probably the less obvious offensive posture that people don't think about is just around discovery and disclosure. Some of the statistics I’ve read indicate that more than 90 percent of the compromises are actually reported externally, rather than the company discovering it.
It's a PR and regulatory nightmare, when someone comes to us and says, "You've been attacked or breached," versus us discovering something and reporting it. Some of the examples I gave were technology, but some of it is just planning and making sure that we’re proactive and report and disclose, rather than seeing it in the headlines.
Gardner: Are there any particular types of technology that help contain an intruder or some other breach?
Rahbari: Some of it is just process, and some of it is technology. When most companies discover a breach, they take people who are already in a full-time job function related to security and put them on the team to investigate. These people don't have the investigative skills and knowledge to deal with incident management, which is truly a specialized science now.
Companies that have experienced breaches and now know how important this is have implemented best practices by having a dedicated team to plan and handle breaches. It's like assigning a SWAT team to a hostage situation versus a police patrol officer.
A SWAT team is trained to handle hostage situations. They're equipped for that. They have a machine gun and sniper rifle instead of a handgun, and they don't have another full-time job to worry about while they’re trying to deal with the crisis. So, from a process and team perspective, that's the first place I would start.
This is even more critical for targeted attacks, which are pretty common nowadays. You have to have the necessary infrastructure to prepare for an event or an incident ahead of time. During the attack, we usually don't disrupt the attacker, or alert them that we’re about to go remedy something. After the remediation is prepared, then we implement that and, in that process, learn about what the attacker is doing.
Some of the steps I recommend, once you know a breach has occurred, is to first pull the entire network off the internet until remediation is complete, blocking the known attacker, domains, and IP addresses. Other simple things include such things as changing compromised passwords.
The preparation ahead of time is important ... because the architecture is the fundamental way in which we can secure our environments.
Some of the steps that any good company should take is first make sure that their networks are separated and isolated. You need a long term network architecture and strategic plan. You also need to establish security zones to separate high risk domains, and make sure you have standards to govern the level of trust between sites and your networks, based on your business requirements.
As far as domain segregation, make sure you do things such as separating Active Directory domains with credentials for your production environments, versus your quality assurance developments and other employee-access environments.
From a trends perspective, there are a number of things that had really helped. Virtualization is one of them. It's a key technique to create segregation between applications.
Gardner: How can organizations get started on this? It looks like an awful lot to go after at once. Is there a path to this—a crawl, walk, run approach—that you would recommend in terms of improving your posture, when it comes to security and containment?
Rahbari: This is a very complicated and difficult thing to learn, and that's where partners and other firms really can be a tremendous help. First, I'd start with an existing organization. Five years ago, it was difficult to sell security to business leaders. Now, those same business leaders are seeing in the paper everyday—the numbers are astronomical—Sony, Google, and others who had breaches.
From the inside, good indication of a security culture change is when you have a dedicated chief information security officer (CISO), the company has a security or risk committee, security is a budget line in the item and not just buried within an IT budget, and security is the business issue.
With the incredible amount of regulatory burden, scrutiny, and oversight, a breach can really tank a company overnight. You read in the paper two months ago, we had a company that lost half of its stock value overnight, after a breach, after Visa was hinting that they might stop using them.
I'd highly recommend that you hire a reputable company to get started, if your particular firm cannot afford to invest in hiring the experts. There are lots of firms that can come. You can outsource to start with, and then as you feel comfortable, bring it in-house and leverage the expertise of this highly, highly specialized field to protect your company and assets.
We automatically stop accepting comments 180 days after a post is published. If you would like to know more about this subject, please contact us and we'll try to help.
Published by: electronicdawn Ltd.