Home on the Palo Alto Networks Cyber Range

You’ve probably heard many horror stories by now about the crazy interviews that companies in Silicon Valley put you though. Sure, some of the questions are downright silly. How would I know how to weigh the moon? But the most insidious are the ones designed to look like skills tests. You may have to spend an hour optimizing a bubble sort or writing some crazy code that honestly won’t have much impact on the outcome of what you’ll be doing for the company.

Practical skills tests have always been the joy and the bane of people the world over. Many disciplines require you to have a practical examination before you can be certified. Doctors are one. The Cisco CCIE is probably the most well-known in IT. But what is the test really quizzing you on? Most people will admit that the CCIE is an imperfect representation of a network at best. It’s a test designed to get people to think about networks in different ways. But what about other disciplines? What about the ones where time is even more of the essence than it was in CCIE lab?

Red Team Go!

I was at Palo Alto Networks Ignite19 this past week and I got a chance to sit down with Pamela Warren. She’s the Director of Government and Industry Initiatives at Palo Alto Networks. She and her team have built a very interesting concept that I loved to see in action. They call it the Cyber Range.

The idea is simple enough on the surface. You take a classroom setting with some workstations and some security devices racked up in the back. You have your students log into a dashboard to a sandbox environment. Then you have your instructors at the front start throwing everything they can at the students. And you see how they respond.

The idea for the Cyber Range came out of military exercises that NATO used to run for their members. They wanted to teach their cyberwarfare people how to stop sophisticated attacks and see what their skill levels were with regards to stopping the people that could do potential harm to nation state infrastructure or worse to critical military assets during a war. Palo Alto Networks get involved in helping years ago and Pamela grew the idea into something that could be offered as a class.

Cyber Range has a couple of different levels of interaction. Level 1 is basic stuff. It’s designed to teach people how to respond to incidents and stop common exploits from happening. The students play the role of a security operations team member from a fictitious company that’s having a very bad week. You learn how to see the log files, collect forensics data, and ultimately how to identify and stop attackers across a wide range of exploits.

If Level 1 is the undergrad work, Cyber Range Level 2 is postgrad in spades. You dig into some very specific and complicated exploits, some of which have only recently been discovered. During my visit the instructors were teaching everyone about the exploits used by OilRig, a persistent group of criminals that love to steal data through things like DNS exfiltration tunnels. Level 2 of the Cyber Range takes you deep down the rabbit hole to see inside specific attacks and learn how to combat them. It’s a great way to keep up with current trends in malware and exploitive behavior.

Putting Your Money Where Your Firewall Is

To me, the most impressive part of this whole endeavor is how Palo Alto Networks realizes that security isn’t just about sitting back and watching an alert screen. It’s about knowing how to recognize the signs that something isn’t right. And it’s about putting an action plan into place as soon as that happens.

We talk a lot about automation of alerts and automated incident response. But at the end of the day we still need a human being to take a look at the information and make a decision. We can winnow that decision down to a simple Yes or No with all the software in the world but we need a brain doing the hard work after the automation and data analytics pieces give you all the information they can find.

More importantly, this kind of pressure cooker testing is a great way to learn how to spot the important things without failing in reality. Sure, we’ve heard all the horror stories about CCIE candidates that typed in debug IP packet detail on core switch in production and watched it melt down. But what about watching an attacker recon your entire enterprise and start exfiltrating data. And you being unable to stop them because you either don’t recognize the attack vector or you don’t know where to find the right info to lock everything down? That’s the value of training like the Cyber Range.

The best part for me? Palo Alto Networks will bring a Cyber Range to your facility to do the experience for your group! There are details on the page above about how to set this up, but I got a great pic of everything that’s involved here (sans tables to sit at):

How can you turn down something like this? I would have loved to put something like this on for some of my education customers back in the day!

Tom’s Take

I really wish I would have had something like the Cyber Range for myself back when I was fighting virus outbreaks and trying to tame Conficker infections. Because having a sandbox to test myself against scripted scenarios with variations run by live people beats watching a video about how to “easily” fix a problem you may never see in that form. I applaud Palo Alto Networks for their approach to teaching security to folks and I can’t wait to see how Pamela grows the Cyber Range program!

For more information about Palo Alto Networks and Cyber Range, make sure to visit http://Paloaltonetworks.com/CyberRange/


Increasing Entropy with Crypto4A

Have you ever thought about the increasing disorder in your life? Sure, it may seem like things are constantly getting crazier every time you turn around, but did you know that entropy is always increasing in the universe? It’s a Law of Thermodynamics!

The idea that organized systems want to fall into disorder isn’t too strange when you think about it. Maintaining order takes a lot of effort and disorder is pretty easy to accomplish by just giving up. Anyone with a teenager knows that the amount of disorder that can be accomplished in a bedroom is pretty impressive.

One place where we don’t actually see a lot of disorder is in the computing realm. Computers are based on the idea that there is order and rationality in everything that we do. This is so prevalent that finding a way to be random is actually pretty hard. Computer programmers have tried a number of ways to come up with random number generators that take a variety of inputs into the formula and come up with something that looks sufficiently random. For most people just wanting the system to guess a number between 1 and 100 it’s not too bad. But when it comes to really, really large numbers like the ones used in cryptography, those pseudorandom numbers aren’t good enough.

This All Looks So Familiar…

One of the reasons for this comes down to good old fashioned efficiency. In the old days computers programmers could rely on people to generate pseudorandom input. By sampling mouse clicks or delay between computer keyboard keystrokes you could easily come up with a number that looks nice and random. However, we’ve taken people out of the loop now. Thanks to the cloud and automation and any one of a number of new ways to reduce human input we’ve managed to remove mouse clicks and keystrokes.

That’s fine for running scripts and programs. It’s even good for building things at a huge scale. But it’s really bad when you need something that looks relatively random. And it’s really, really bad when your program relies on that randomness to keep you secure. Kind of like key generation in Public Key Cryptography (PKI).

A group of security researchers working for the National Institute of Standards and Technology (NIST) found out a few years ago that public keys were starting to collide at greater rates than random chance. The study, conducted in 2012, found that 5% of HTTPS and 10% of SSH public keys were duplicates. A collision in a hashing algorithm is when two inputs produce the same output, which renders that hashing function broken. In PKI, having a two different inputs output the same public key is really bad, because it could lead to key collisions that impact a variety of service.

What caused it? As it turns out, lack of orderly disorder. Because automation and non-human interaction have led to other pseudorandom inputs being used in key generation it appeared to the researchers that the same inputs were being used all over the place. That meant that a lot of the public keys that were being generated were being done in such as way as to make collisions more likely. When you look at how many things are relying on automated sources to generate keys it can be quite scary. Think about a smart lightbulb or other IoT device that’s trying to generate pseudorandom input from a CPU that’s just big enough to turn things on. Now imagine that CPU multiplied by the number of smart lightbulbs out there. Not a pleasant thought, is it?

Disorder In The Court

This fascinating discussion came from an interview I had with Bruno Couillard, the President and CTO of Crypto4A. Crypto4A is a company that provides Entropy-as-a-Service. What exactly does that mean?

Crypto4A has an appliance they call QAOS. QAOS is designed to give you the best possible disorder that you can get. It does this the old fashioned way. Instead of trying to use software as a Random Number Generator (RNG) QAOS instead uses hardware sources to generate entropy for their RNG. This includes a quantum RNG, which produces high quality disorder that’s difficult to fake any other way.

QAOS is designed to feed software with entropy to generate randomness sufficient to prevent PKI public key collisions. The software developers can follow the NIST guidelines on EaaS to have the program call an entropy source. QAOS, acting as that entropy source, will seed the RNG on the target system with good randomness and allow it to generate good keys. This could also be configured in the kernel of the OS to call a system like QAOS on boot and start the seed value with a good amount of random entropy in the case of old programs that can’t be modified to call anything other than a system-based RNG source like /dev/random/.

Tom’s Take

The NIST guidelines around EaaS are constantly evolving, but the idea that companies are already racing to fill the void that has been created by insufficient randomness in cryptography is telling. When you think about nth the number of devices that are going to be using PKI for secure communications, the need for something like Crypto4A QAOS is pretty clear. If we are going to rely on automated systems to run our daily lives, we need to have the resources in place to ensure they have a solid foundation of randomness to build on.

What Makes IoT A Security Risk?

IoT security is a pretty hot topic in today’s world. That’s because the increasing number of smart devices is causing issues with security professionals everywhere. Consumer IoT devices are expected to top 20 billion by 2020. And each of these smart devices represents an attack surface. Or does it?

Hello, Dave

Adding intelligence to a device increases the number of ways that it can compromised. Take a simple thermostat, for example. The most basic themostat is about as dumb as you can get. It uses the expansion properties of metal to trigger switches inside of the housing. You set a dial or a switch and it takes care of the rest. Once you start adding things like programmability or cloud connection, you increase the number of ways that you can access the device. Maybe it’s a webpage or an app. Maybe you can access it via wireless or Bluetooth. No matter how you do it, it’s more available than the simple version of the thermostat.

What about industrial IoT devices? The same rule applies. In this case, we’re often adding remote access to Supervisory Control And Data Acquistion (SCADA) systems. There’s a big market from enterprise IT providers to create secured equipment that allows access to existing industrial equipment from centralized control dashboards. It makes these devices “smart” and allows you to make them easier to manage.

Industrial IoT has the same kind of issues that consumer devices do. We’re increasing the number of access avenues to these devices. But does that mean they’re a security risk? The question could be as simple as asking if the devices are any easier to hack than their dumb counterparts. If that is our only yardstick, then the answer is most assuredly yes they are a security risk. My fridge does not have the ability for me to access it over the internet. By installing an operating system and connecting it to the wireless network in my house I’m increasing the attack surface.

Another good example of this increasing attack surface is in home devices that aren’t consumer focused. Let’s take a look at the electrical grid. Our homes are slowly being upgraded with so-called “smart” electrical meters that allow us to have more control over power usage in our homes. It also allows the electric companies to monitor the devices more closely and read the electric meters remotely instead of needing to dispatch humans to read those meters. These smart meters often operate on Wi-Fi networks for ease-of-access. If all we do is add the meters to a wireless network, are we really creating security issues?

Bigfoot-Sized Footprints

No matter how intelligent the device, increasing access avenues to the device creates security access issues. A good example of this is the “hidden” diagnostic port on the original Apple Watch. Even though the port had no real use beyond internal diagnostics at Apple, it was a tempting target for people to try and get access to the system. Sometimes these hidden ports can dump hidden data or give low-level access to areas of the system that aren’t normally available. While the Apple Watch port didn’t have this kind of access, other devices can offer it.

Giving access to any device allows you to attack it in a way that can gain you access that can get you into data that you’re not supposed to have. Sure, a smart speaker is a very simple device. But what if someone found a way to remotely access the data and capture the data stream? Or the recording buffer? Most smart speakers are monitoring audio data listening for their trigger word to take commands. Normally this data stream is dumped. But what if someone found a way to reconstruct it? Do you think that could qualify as a hack? All it takes is an enterprising person to figure out how to get low-level access. And before you say it’s impossible, remember that we allow access to these devices in other ways. It’s only a matter of time before someone finds a hole.

As for industrial machines, these are even more tempting. By gaining access to the master control systems, you can cause some pretty credible havoc with their programming. You can shut down all manner of industrial devices. Stuxnet was a great example of writing a very specific piece of malware that was designed to cause problems for a specific kind of industrial equipment. Because of the nature of the program it was very difficult to figure out exactly what was causing the issues. All it took was access to the systems, which was reportedly caused by hiding the program on USB drives and seeding them in parking lots where they would be picked up and installed in the target facilities.

IoT devices, whether consumer or enterprise, represent potential threat vectors. You can’t simply assume that a simple device is safe because there isn’t much to hack. The Mirai bonnet exploited bad password hygiene in devices to allow them to be easily hacked. It wasn’t a complicated silicon-level hack or a coordinated nation state effort. It was the result of someone cracking a hard-coded password and exploiting that for their own needs. Smart devices can be made to make dumb decisions when used improperly.

Tom’s Take

IoT security is both simple and hard at the same time. Securing these devices is a priority for your organization. You may never have the compromised, but you have to treat them just like you would any other device that could potentially be hacked and turned against you. Zero-trust security models are a great way to account for this, but you need to make sure you’re not overlooking IoT when you build that model. Because the invisible devices helping us get our daily work done could quickly become the vector for hacking attacks that bring our day to a grinding halt.

Facebook’s Mattress Problem with Privacy

If you haven’t had a chance to watch the latest episode of the Gestalt IT Rundown that I do with my co-workers every Wednesday, make sure you check this one out. Because it’s the end of the year it’s customary to do all kinds of fun wrap up stories. This episode focused on what we all thought was the biggest story of the year. For me, it was the way that Facebook completely trashed our privacy. And worse yet, I don’t see a way for this to get resolved any time soon. Because of the difference between assets and liabilities.

Contact The Asset

It’s no secret that Facebook knows a ton about us. We tell it all kinds of things every day we’re logged into the platform. We fill out our user profiles with all kinds of interesting details. We click Like buttons everywhere, including the one for the Gestalt IT Rundown. Facebook then keeps all the data somewhere.

But Facebook is collecting more data than that. They track where our mouse cursors are in the desktop when we’re logged in. They track the amount of time we spend with the mobile app open. They track information in the background. And they collect all of this secret data and they store it somewhere as well.

This data allows them to build an amazingly accurate picture of who we are. And that kind of picture is extremely valuable to the right people. At first, I thought it might be the advertisers that crave this kind of data. Advertisers are the people that want to know exactly who is watching their programs. The more data they have about demographics the better they can tailor the message. We’ve already seen that with specific kinds of targeted posts on Facebook.

But the people that really salivate over this kind of data live in the shadows. They look at the data as a way to offer new kinds of services. Don’t just sell people things. Make them think differently. Change their opinions about products or ideas without them even realizing it. The really dark and twisted stuff. Like propaganda on a whole new scale. Enabled by the fact that we have all the data we could ever want on someone without even needing to steal it from them.

The problem with Facebook collecting all this data about us is that it’s an asset. It’s not too dissimilar from an older person keeping all their money under a mattress. We scoff at that person because a mattress is a terrible place to keep money. It’s not safe. And a bank will pay you keep your money there, right?

On the flip side, depending on the age of that person, they may not believe that banks are safe. Before FDIC, there was no guarantee your money would be repaid in a pinch. And if the bank goes out of business you can’t get your investment back. For a person that lived through the Great Depression that had to endure bank holidays and the like, keeping your asset under a mattress is way safer than giving it to someone else.

As an aside here, remember that banks don’t like leaving your money laying around either. If you deposit money in a bank, they take that money and invest it in other places. They put the money to work for them making money. The interest that you get paid for savings accounts and the like is just a small bonus to encourage you to keep your money in the bank and not to pull it out. That’s why they even have big disclaimers saying that your money may not be available to withdraw at a moment’s notice. Because if you do decide to get all of your money out of the bank at once, they need to go find the money to give you.

Now, let’s examine our data. Or, at least the data that Facebook has been storing on us. How do you think Facebook looks at that data? Do you believe they want to keep it under the mattress where it’s safe from the outside world? Do you think that Facebook wants to keep all these information locked in a vault somewhere where no one can get to it?

Or perhaps Facebook looks at your data as an asset like a bank does. Instead of keeping it around and letting it sit fallow they’d rather put it to work. That’s the nature of a valuable asset. To the average person, their privacy is one of the most important parts of their lives. To Facebook, your privacy is simply an asset. It can either sit by itself and make them nothing. Or it can be put to use by Facebook or third-party companies to make more money from the things that they can do with good data sources. To believe that a company like Facebook has your best interests at heart when it comes to privacy is not a good bet to make.

Would I Lie-ability To You?

In fact, the only thing that can make Facebook really sit up and pay attention is if that asset they have farmed out and working for them were to suddenly become a liability for some reason. Liabilities are a problem for companies because they are the exact opposite of making money. They cost money. Just as the grandmother in the above example sees an insolvent bank as a liability, so too would someone see a bad asset as a possible exposure.

Liabilities are a problem. Anything that can be an exposure is an issue for company, especially one with investors that like to get dividends. Any reduction in profit equals a loss. Liabilities on a balance sheet are giant red flags for anyone taking a close look at the operations of a business.

Turning Facebook’s data assets into a liability is the only way to make them sit up and realize that what they’re doing is wrong. Selling access to our data to anyone that wants it is a horrible idea. But it won’t stop until there is some way to make them pay through he nose for screwing up. Up until this year, that was a long shot at best. Most fines were in the thousands of dollars range, whereas most companies would pay millions for access to data. A carefully crafted statement admitting no fault after the exposure was uncovered means that Facebook and the offending company get away without a black mark and get to pocket all their gains.

The European GDPR law is a great step in the right direction. It clearly spells out what has to happen to keep a person’s data safe. That eliminates wiggle room in the laws. It also puts a stiff fine in place to ensure that any violations can be compounded quickly to drain a company and turn data into a liability instead of an asset. There are moves in the US to introduce legislation similar to GDPR, either at the federal level or in individual states like California, the location of Facebook’s headquarters.

That’s not to say that these laws are going to get it right every time. There are people out there that live to find ways to turn liabilities into assets. They want to find ways around the laws and make it so that they can continue to take their assets and make money from them even if the possibility of exposure is high. It’s one thing when that exposure is the money of people that invested in them. It’s another thing entirely when it’s personally identifiable information (PII) or protected information about people. We’re not imaginary money. We live and breath and exist long past losses. And trying to get our life back on track after an exposure is not easy for sure.

Tom’s Take

If I sound grumpy, it’s because I am tired of this mess. When I was researching my discussion for the Gestalt IT Rundown I simply Googled “Facebook data breach 2018” looking for examples that weren’t Cambridge Analytica. The number was more than it should have been. We cry about Target and Equifax and many other exposures that have happened in the last five years, but we also punish those companies by not doing business with them or moving our information elsewhere. Facebook has everyone hooked. We share photos on Facebook. We RSVP to events on Facebook. And we talk to people on Facebook as much or more than we do on the phone. That kind of reach requires a company to be more careful with who has access to our data. And if the solution is building the world’s biggest mattress to keep it all safe put me down for a set of box springs.


Some Random Thoughts From Security Field Day

I’m spending the week in some great company at Security Field Day with awesome people. They’re really making me think about security in some different ways. Between our conversations going to the presentations and the discussions we’re having after hours, I’m starting to see some things that I didn’t notice before.

  • Security is a hard thing to get into because it’s so different everywhere. Where everyone just sees one big security community, it is in fact a large collection of small communities. Thinking that there is just one security community would be much more like thinking enterprise networking, wireless networking, and service provider networking are the same space. They may all deal with packets flying across the wires but they are very different under the hood. Security is a lot of various communities with the name in common.
  • Security isn’t about tools. It’s not about software or hardware or a product you can buy. It’s about thinking differently. It’s about looking at the world through a different lens. How to protect something. How to attack something. How to figure all of that out. That’s not something you learn from a book or a course. It’s a way of adjusting your thinking to look at problems in a different way. It’s not unlike being in an escape room. Don’t look at the objects like you normally would. Instead, think about them with unique combinations that get you somewhere different than where you thought you needed to be.
  • Security is one of the only IT disciplines where failure is an acceptable outcome. If we can’t install a router or a wireless access point, it’s a bad job. However, in security if you fail to access something that should have been secured it was a success. That can lead to some very interesting situations that you can find yourself in. It’s important to realize that you also have to properly document your “failure” so people know what you tried to do to get there. Otherwise your success may just be a lack of proper failure.

Tom’s Take

I’m going to have some more thoughts from Security Field Day coming up another time. There’s just too much to digest at one time. Stay tuned for some more great discussions and highlights of my first real foray in the security community!

What Makes a Security Company?

When you think of a “security” company, what comes to mind? Is it a software house making leaps in technology to save us from DDoS attacks or malicious actors? Maybe it’s a company that makes firewalls or intrusion detection systems that stand guard to keep the bad people out of places they aren’t supposed to be. Or maybe it’s something else entirely.

Tradition Since Twenty Minutes Ago

What comes to mind when you think of a traditional security company? What kinds of technology do they make? Maybe it’s a firewall. Maybe it’s an anti-virus program. Or maybe it’s something else that you’ve never thought of.

Is a lock company like Schlage a security company? Perhaps they aren’t a “traditional” IT security company but you can guarantee that you’ve seen their products protecting data centers and IDF closets. What about a Halon system manufacturer? They may not be a first thought for security, but you can believe that a fire in your data center is going cause security issues. Also, I remember that I learned more about Halon and wet/dry pipe fire sprinkler systems from my CISSP study than anywhere else.

The problem with classifying security companies as “traditional” or “non-traditional” is that it doesn’t reflect the ways that security can move and change over the course of time. Even for something as cut-and-dried as anti-virus, tradition doesn’t mean a lot. Symantec is a traditional AV vendor according to most people. But the product that used to be called Norton Antivirus and the product suite that now includes is are worlds apart in functionality. Even though Symantec is “traditional”, what they do isn’t. And when you look at companies that are doing more advanced threat protection mechanisms like deception-based security or using AI and ML to detect patterns, the lines blur considerably.

But, it doesn’t obviate the fact that Symantec is a security company. Likewise, a company can be a security company even if they security isn’t their main focus. Like the Schlage example above, you can have security aspects to your business model without being totally and completely focused on security. And there’s no bigger example of this than a company like Cisco.

A Bridge Not Far Enough?

Cisco is a networking company right? Or are they a server company now? Maybe they’re a wireless company? Or do they do cloud now? There are many aspects to their business models, but very few people think of them as a security company. Even though they have firewalls, identity management, mobile security, Malware protection, VPN products, Email and Web Security, DNS Protection, and even Threat Detection. Does that mean they aren’t really a security company?

It could be rightfully pointed out that Cisco isn’t a security company because many of these technologies they have were purchased over the years from other companies. But does that mean that their solutions aren’t useful or maintained? As I was a doing research for this point, a friend pointed out the story of Cisco MARS and how it was purchased and ultimately retired by Cisco. However, the Cisco acquisition of Protego that netted them MARS happened in 2004. The EOL announcement was in 2011, and the final end-of-support was in 2016. Twelve years is a pretty decent lifetime for any security product.

The other argument is that Cisco doesn’t have a solid security portfolio because they have trouble integrating their products together. A common criticism of large companies like Cisco or Dell EMC is that it is too difficult to integrate their products together. This is especially true in situations where the technologies were acquired over time, just like Cisco.

However, is the converse true? Are standalone products easier to integrate? Is is more simple to take solutions from six different companies and integrate them together in some fashion? I’d be willing to be that outside of robust API support, most people will find that integrating security products from different vendors is as difficult (if not more so) than integrating products from one vendor. Does Cisco have a perfect integration solution? No, they don’t. But why should they? Why should it be expected that companies that acquire solutions immediate burn cycles to make everything integrate seamlessly. Sure, that’s on the roadmap. But integrations with other products is on everyone’s road map.

The last argument that I heard in my research is that Cisco isn’t a security company because they don’t focus on it. They’re a networking (or wireless or server) company. Yet, when you look at the number of people that Cisco has working in a specific business unit on a product, it can often be higher headcount that some independent firms have working on their solutions. Does that mean that Cisco doesn’t know what they’re doing? Or does it mean that individual organizations can have multiple focuses? That’s a question for the customers to answer.

Tom’s Take

I take issue with a definition of “traditional” versus non-traditional. For the reason that Apple is a traditional computer company and so is Wang Computers. Guess which one is still making computers? And even in the case of Apple, you could argue that their main line-of-business is mobile devices now. But, does anyone dispute Apple’s ability to make a laptop? Would a company that does nothing but make laptops be a “better” computer company? The trap of labels like that is that it ignores a significant amount of investment in business at the expense of a quick and easy label. What makes a company a computer company or a security company isn’t how they label themselves. It’s what they do with the technology they have.

Security Is Bananas

I think we’ve reached peak bombshell report discussion at this point. It all started this time around with the big news from Bloomberg that China implanted spy chips into SuperMicro boards in the assembly phase. Then came the denials from Amazon and Apple and event SuperMicro. Then started the armchair quarterbacking from everyone, including TechCrunch. From bad sources to lack of technical details all the way up to the crazy conspiracy theories that someone at Bloomberg was trying to goose their quarterly bonus with a short sale or that the Chinese planted the story to cover up future hacking incidents, I think we’ve covered the entire gamut of everything that the SuperMicro story could and couldn’t be.

So what more could there be to say about this? Well, nothing about SuperMicro specifically. But there’s a lot to say about the fact that we were both oblivious and completely unsurprised about an attack on the supply chain of a manufacturer. While the story moved the stock markets pretty effectively for a few days, none of the security people I’ve talked to were shocked by the idea of someone with the power of a nation state inserting themselves into the supply chain to gain the kind of advantage needed to execute a plan of collection of data. And before you scoff, remember we’re only four years removed from the allegation that the NSA had Cisco put backdoors into IOS.

Why are we not surprised by this idea? Well, for one because security is getting much, much better at what it’s supposed to be doing. You can tell that because the attacks are getting more and more sophisticated. We’ve gone from 419 scam emails being deliberately bad to snare the lowest common denominator to phishing attacks that fool some of the best and brightest out there thanks to a combination of assets and DNS registrations that pass the initial sniff test. Criminals have had to up their game because we’re teaching people how to get better at spotting the fakes.

Likewise, technology is getting better at nabbing things before we even see them. Take the example of Forcepoint. I first found out about them at RSA this year. They have a great data loss prevention (DLP) solution that keeps you from doing silly things like emailing out Social Security Numbers or credit card information that would violate PCI standards. But they also have an AI-powered analysis engine that is constantly watching for behavioral threats. If someone does this on accident once it could just be a mistake. But a repeated pattern of behavior could indicate a serious training issue or even a malicious actor.

Forcepoint is in a category of solutions that are making the infrastructure smarter so we don’t have to be as vigilant. Sure, we’re getting much better at spotting things to don’t look right. But we also have a lot of help from our services. When Google can automatically filter spam and then tag presented messages as potentially phishing (proceed with caution), it helps me start my first read through as a skeptic. I don’t have to exhaust my vigilance for every email that comes across the wire.

The Dark Side Grows Powerful Too

Just because the infrastructure is getting smarter doesn’t mean we’re on the road to recovery. It means the bad actors are now exploring new vectors for their trade. Instead of 419 or phishing emails they’re installing malware on systems to capture keystrokes. iOS 12 now has protection from fake software keyboards that could capture information when something is trying to act as a keyboard on-screen. That’s a pretty impressive low-level hack when you think about it.

Now, let’s extrapolate the idea that the bad actors are getting smarter. They’re also contending with more data being pushed to cloud providers like Amazon and Azure. People aren’t storing data on their local devices. It’s all being pushed around in Virginia and Oregon data centers. So how do you get to that data? You can’t install bad software on a switch or even a class of switches or even a single vendor, since most companies are buying from multiple vendors now or even looking to build their own networking stacks, ala Facebook.

If you can’t compromise the equipment at the point of resale, you have to get to it before it gets into the supply chain. That’s why the SuperMicro story makes sense in most people’s heads, even if it does end up not being 100% true. By getting to the silicon manufacturer you have a entry point into anything they make. Could you imagine if this was Accton or Quanta instead of SuperMicro? If there was a chip inside every whitebox switch made in the last three years? If that chip had been scanning for data or relaying information out-of-band to a nefarious third-party? Now you see why supply chain compromises are so horrible in their potential scope.

This Is Bananas

Can it be fixed? That’s a good question that doesn’t have a clear answer. I look at it like the problem with the Cavendish banana. The Cavendish is the primary variant of the banana in the world right now. But it wasn’t always that way. The Gros Michel used to be the most popular all the way into the 1950s. It stopped because of a disease that infected the Gros Michel and caused entire crops to rot and die. That could happen because bananas are not grown through traditional reproductive methods like other crops. Instead, they are grafted from tree to tree. In a way, that makes almost all bananas clones of each other. And if a disease affects one of them, it affects them all. And there are reports that the Cavendish is starting to show signs of a fungus that could wipe them out.

How does this story about bananas relate to security? Well, if you can’t stop bananas from growing everywhere, you need to take them on at the source. And if you can get into the source, you can infect them without hope of removal. Likewise, if you can get into the supply chain and start stealing or manipulating data a low level, you don’t need to worry about all the crazy protections put in at higher layers. You’ll just bypass them all and get what you want.

Tom’s Take

I’m not sold on the Bloomberg bombshell about SuperMicro. The vehement denials from Apple and Amazon make this a more complex issue than we may be able to solve in the next couple of years. But now that the genie is out the bottle, we’re going to start seeing more and more complicated methods of attacking the merchant manufacturers at the source instead of trying to get at them further down the road. Maybe it’s malware that’s installed out-of-the-box thanks to a staging server getting compromised. Maybe it’s a hard-coded backdoor like the Xiamoi one that allowed webcams to become DDoS vectors. We can keep building bigger and better protections, but eventually we need to realize that we’re only one threat away from extinction, just like the banana.