Acreto Security

Russian Hacker Caught and Convicted: From US With Love.

Written by Babak Pasdar, CEO and CTO of Acreto.

A little while ago, a client called me in to do a security operations ‘best practices’ education session. They were a dot com site that had recently spun off from one of the major financials. They had not yet laid down their sec ops roots and were still engaged in establishing the fundamentals. They wanted an informal education session to get the entire team on the same page.

Their conference room was packed with their security team as well as several people from their operations center, which I had requested. In many instances, the ops team is on the front line and often identifies and conducts the initial steps in handling security incidents.

At some point during the session, I started to talk about scammers. One trick that malicious people use is to acquire domain names that are similar to the site they are targeting. Since the client was a financial and their site contained personal information for hundreds of thousands of consumers, and was an attractive target. I first recommended they acquire or actively monitor all sound-alike and similar domains. For example, if their domain name is jacks.com, spelled J A C K S, they should acquire or monitor the domains spelled J A X dot com and J A K S dot com.

Second, I recommended that all permutations of domains that could be mis-typed by users should be acquired or monitored as well; specifically, any combination of surrounding characters on the keyboard for each letter that makes up their domain name. For example, if their domain name is abc.com, they should monitor domains where the ‘A’ in abc.com is replaced with S, X, Z, W, and Q. If a company wanted to take it a step further, they would cover the immediate two surrounding characters on the keyboard as well. Should users mistype, which they often do, they should not be directed to a look-alike site that they would innocently offer their credentials.

Third, I suggested that the plural version of the words included in their domain name should be acquired and monitored. As I was making this third point, I typed in the plural of their domain name – and their site showed up. I thought I had made a typo, that through muscle memory I had entered in their correct domain name. I double checked, and I had typed exactly what I intended to type – the incorrect, plural variant.

I was impressed. I thought to myself that they were ahead of me and had already acquired the plural domain and redirected it to their site. “Smart! You guys already got this?” I said to the group. I looked around the room and saw confused expressions all around. Finally, someone said, “I don’t think that we did – I’m pretty sure we didn’t.”

After a Dig on the Fully Qualified Domain Name (FQDN) and an MTR (a better traceroute) it became clear that the site was not theirs. It looked exactly like their site including the login page. However, it was not using their IP block nor any of their ISPs. It traced back to Las Vegas, Nevada.

Needless to say, the training session abruptly ended and became a real-life incident response. The organization’s executives, their general counsel, all security team members, and all IT managers and above joined an emergency meeting in the conference room. Anyone not on-site joined via conference bridge.

During the meeting, their sharp help-desk manager offered that he had seen an increase in the number of calls for password reset requests in the past two weeks. We started connecting the dots.
We came away from the meeting with several action items:

• We needed to determine if there was a compromise, and if so, how many users it impacted and its duration.
• The help-desk team set out to cross correlate password reset support calls and the date/time of failed authentication logins in their logs.
• They would identify any users who called for a password reset whom had no corresponding failed login attempts in the logs. There was roughly a dozen dating back only two weeks.
• The help-desk team contacted these users and established completely new identities for them.
• My team was to implement an emergency infrastructure should the malicious person attempt to use the stolen identities.
• I reached out to my contacts in the FBI cyber-crime team and reported the issue, and Agent Brown from the New York cybercrimes team was assigned to our case.
• We contacted a law firm with experience in cyber crimes along with the organization’s retained counsel.
• The legal team started to outline a notice as was required by compliance in preparation, should notifications be necessary.

After this, my team members and I set out to execute on a plan to identify and catch the person.

First, a honeypot. The compromised user credentials correlated by the helpdesk were redirected to a training system that looked and functioned just like their application, but contained dummy data. With this in place, the risk that any (more?) data theft, manipulation or deletion was mitigated.

Then, we implemented a high performance packet capture system using a powerful server, hardware offloading network interface and several open-source tools to collect all communications from the malicious person/people. We made sure that the packet capture system was implemented and processed with proper evidentiary chain of custody standards.

Finally, we configured the units to send us text messages as soon as any of the compromised accounts were accessed.

We were finally ready to track the malicious people.

In less than forty-eight hours we architected, acquired the highly specialized equipment required, and configured and tested the infrastructure. I then set out to document everything, including the operations runbook for these new systems, which included evidentiary chain of custody handling of any evidence collected.

I personally spent near seventy-two hours straight at the customer’s data center hopped up on adrenaline and coffee. It’s rare to catch hackers and scammers, and I felt strongly that we had a good chance of doing so in this case.

In the meantime, the FBI requested and received a subpoena for the IP address of the server as well as the domain name registrar. Fortunately, the ISP provided the physical address associated with the identified IP address quickly.

Agent Brown called the FBI field office in Nevada and requested agents drive by and visualize the address location. A few hours later we received information that the address was actually a car dealership. The FBI agents in Nevada managed to trace the ISP connection to the basement of the dealership. When they inquired about the Internet connection, the dealership informed them that the basement was rented to another party who was hardly ever there.

Technically, the malicious people had not done anything substantially criminal. So between the customer, the FBI and my team we decided to hang back and wait for the malicious people to attempt access to the customer system, and more importantly, to download personal identity information. There was no risk to any of the site users since the data the malicious people would access was made up training data.

We didn’t have to wait long. At 3:00am early morning the following day my phone started buzzing with alerts. I quickly logged on to see what had transpired. Jackpot! The malicious people had logged on under three different accounts and had systematically accessed multiple identifies before generating a report that can only be identified as an identity theft starter kit.

A quick check showed a Canadian IP address as the source. Every packet of the communications was collected and logged. We had all that was required to completely recreate and replay the malicious people’s entire effort.

The session was short. It had only lasted 15 minutes. But it was all that was necessary. There were no other attempts that day.

Early the following morning, we contacted Agent Brown and the cybercrime task force supervisor and arranged for collection of the evidence. During the call we also determined our next course of action.
The FBI could have reached out to the Canadian authorities, but thought it best to try to lure the person to the US.

The plan was that the FBI would get a court order to confiscate the computer in Las Vegas. If they spotted cameras they would simply disconnect the Internet connection at the Network Terminal outside the building.

And then – the FBI surprised us. They had a person of interest in the case. They did not share many details about how they found this person of interest. Our best guess is that the person had been on the FBI’s radar, and had somehow been associated with the stolen identity which was used to fraudulently pay for the acquired domain name and the Las Vegas basement housing the computer.

If all was to go as planned, the malicious person would think there is a technical issue and come to fix it.

Later that morning the FBI Agent Brown came to our offices and we held an evidence hand-off ceremony. The next day we noticed that the scam site had gone down. Now there was not much else for us to do but wait.

All was quiet for a while and life started to resume normalcy. Two weeks later we got word that there had been an arrest!

It was a Russian whom a few days after the site had gone down had flown to Canada and from Canada to Las Vegas. He was arrested at the airport port of entry. Apparently, when presented with the evidence he made a plea bargain and soon after plead guilty at the hearing.

The team’s dedication, professionalism and expertise drove this incident’s success. Both the customer and my team operated flawlessly together, and the FBI came through in a big way. At a time when hackers attack indiscriminately, it felt great to catch one and snag a win for the good guys.

Learn more by visiting our web site: Acreto.io — On Twitter: @acretoio and if you haven’t done so, sign up for the Acreto Crypto-n-IoT podcast. You can get it from Apple – Google or your favorite podcast app.

Next up -- read about Russian Nation-State Hackers and What We're Not Doing About It here!

About Acreto IoT Security

Acreto IoT Security delivers advanced security for IoT Ecosystems, from the cloud. IoTs are slated to grow to 50 Billion by 2021. Acreto’s Ecosystem security protects all Clouds, users, applications, and purpose-built IoTs that are unable to defend themselves in-the-wild. The Acreto platform offers simplicity and agility, and is guaranteed to protect IoTs for their entire 8-20 year lifespan. The company is founded and led by an experienced management team, with multiple successful cloud security innovations. Learn more by visiting Acreto IoT Security on the web at acreto.io or on Twitter @acretoio.

The Business of Security vs. Security of Business

Written by Babak Pasdar, CEO and CTO of Acreto.

The security industry has spent a lot of time over the past 30 years thinking of imaginative ways to put lipstick on today's cybersecurity pig.

It's like a one hit wonder band who never adapted, playing the same song and putting on the same show over and over, even though their fans, the industry and the zeitgeist as a whole have evolved and transitioned.

We are more distributed and mobile than ever. Yet the security industry remains unevolved, putting on the same show – playing their all-time favorites like “On-Device Security” and their mega-hit “Gateway Security”. Gateway security is an especially nuanced piece with broad range. There’s the firewall, intrusion prevention, VPN gateway, the proxy, url and content filters, and the component that binds them – SIEM. And that’s the consolidated version of a lengthier and more complicated original score.

Compute has changed and continues to change dramatically in front of our eyes. Clouds, SaaS, Mobile devices and the big daddy of them all – IoT – are contorting traditional security models and tools in ways never intended – until something breaks. And today, everything is breaking since security as we know it dates back to the medieval ages.

Let’s Get Medieval On Security.

The king builds a castle (the network), puts a moat and draw-bridge around it (gateway security) and posts sentries at the gate with special instructions (security policy).

Need to operate outside the castle? If you have the strength (compute resources) and are wealthy enough to afford it (budget), you can put on custom armor (on-device security) and head out as a knight (remote user). Being a knight is exhausting though. Yes, you are well protected, but it burns a lot of energy (security team resources).

However, commoners have to assume risk and live in a state of constant vulnerability. Clouds and IoT have driven the vast majority of our functions and users to operate "outside the castle". In fact, the business of the king’s court is now distributed. Commoners live and work remote, never needing to step foot in the castle.

There are even scenarios where some commoners operate and service other kingdoms near and far. When the court subjects are remote and distributed, the king has two options – insist on keeping the castle, moat and drawbridge or adapt. So far the security industry has bitterly resisted adapting. Why -- Tradition? Lack of alternatives? It's what they know? Or a combination of these.

Gateway security still has its uses, however, the gateway security model is long in the tooth and its use-cases diminishing by the week. And on-device security has been an expensive, ineffective and unsustainable failure. How can you package up an entire data center's worth of security functions in a $5 sensor with the compute resources of a Timex watch.

What the cloud started, IoTs have finished. In the past compute was network-centric, now it is distributed all over and even mobile. And we like it. Initially CISOs tried to control users by saying no to cloud and SaaS. Users wouldn't have it. They shrugged, walked away, and did it anyway. There was no putting that toothpaste back in the tube once they got a taste of cloud and SaaS.

Compute and technology has been democratized, however the way we secure is still medieval.

We have offered hackers the overwhelming advantage all the while spending billions and billions on security. Vendors continue to monetize on medieval security tools ill-suited to the new dominant compute model. How does this make sense?

There are a few reasons:

First, it's what people know and have bought into. There are 30 plus years of approaches and methods, tools and technologies, processes and performance indicators that have been developed around medieval security. It has become muscle memory for many who spent years honing their skills around these approaches.

Just imagine if suddenly, through magical circumstances, the rule of thumb became NOT to apply pressure to bleeding wounds. The countless developed methods, processes, tools, and even tangential functions like billing would be impacted. The result would be chaos! Arguably security is experiencing a mild form of chaos now.

Second, there are a lot of vendor-centric security professionals that know and understand security through the prism of a particular vendor. This is not meant to be derogatory since these professionals are the backbone of the security industry. However many are not security operators, they are security product managers.

In most instances, along with functional and integration capabilities, security is but one of multiple features that security tools sport. Many security professionals are really, really good at keeping the lights on and packets flowing – and rely on the product do its security stuff.

Some vendors are so big and influential that more security professionals than we like to admit are exclusively committed to their tools. These professionals have done the economic calculus and have built their careers around a single brand, strictly based on market opportunity. Many evolve when vendors say it's time to evolve for job prospect purposes. And the evolution of certain security professionals is curiously bound to the vendor's business strategy. An arrangement that benefits the vendor and the professional – just not security.

This brings me to the third point: the business of security vs. security of business.

It takes many years for new and emerging approaches or technologies to become mainstream. Large influential vendors are focused on squeezing every last bit of economic value from their existing technology investments, while small innovative companies just don’t have the market megaphone. And pay-to-play analyst firms confuse matters further by offering tilted and skewed recommendations.

Now, let’s talk about the Cyber Hare vs. the Security Turtle.

Hackers are cutting-edge. They are imaginative. They formulate crazy ideas meant to break the rules. The security industry counters with security professionals who are compelled to be conservative – to a fault.

Hackers don’t care about function and performance, whereas organizations prioritize both over security. Hackers can experiment and fail countless times, forging their own path along the way, while organizations identify gaps by virtue of emerging product categories. Often it takes anywhere between three to five years, depending on the organization, to implement new product categories for an emerging threat type. At that point the threat is not so emerging anymore!

Moreover, organizations befuddle themselves by implementing a process, a very organized one at that, developed to assure failure. This includes assessing requirements, assigning budget, talking to Gartner to see who paid them most, evaluating several brands, selecting a technology, negotiating legal, purchasing, implementation, integration, administration, management, monitoring and troubleshooting. Where is the agility?!

Aside from the security functions the product offers, nothing in the process above even comes close to security operations.

What does this mean? It means that hackers have a significant upper hand. This upper hand is so overwhelmingly one-sided that it has evolved from having the ability to impact business, to the ability to devastate economies and undermine democracies.

Cyber - The Longest War.

Today, everyone talks about the war in Afghanistan as our longest running conflict. In the near future this distinction will easily be awarded to the global cyber-war. Every day, much like other security professionals, I see this war from our operations center. I see Russia, China, North Korea, Iran and even some allies wage war against our infrastructure. If not by Name (IP Address), then by reputation (APT).

If we have learned anything from the Afghani and Iraqi conflicts it's that success does not always require a standing army. Special Operations have radically shifted the methods of war. Not only is this cheaper and faster, but also more effective to achieve many missions around the world. Today the SpecOps model is being employed in the Syrian conflict.

Maybe we should learn from the military and apply seismic shifts to our security approach. Here's how:

First, let's eliminate products from the equation. Building one-off security using tools that are ill-fitted to address the emerging distributed and mobile compute model is security suicide. Products are always out-of-date and security teams burn valuable resources performing technology refreshes, managing and troubleshooting products rather than operating security.

Security as a utility is a much more effective approach. It is simpler and much faster to sign up and turn on, than to buy and build out! Make implementation easy and let the development, upgrades, updates and keeping the lights on be someone else's problem. The time your team is not spending on babysitting products can be put to better use operating security.

Second, fight hackers with (ethical) hackers. Build or train security teams of operators – not product administrators. Make your team critical thinkers who focus on “how to break things” rather than the mundane keeping the lights on tasks. Not all hackers are foul tempered, tattoo laced, twenty-something rock stars with an ego. There are many agreeable, thoughtful and reliable ethical hackers that can serve in foundational roles on your team. Most importantly, empower them and involve them from the beginning at the application design, development and roll out phases.

The traditional medieval security model is not failing, it has already failed spectacularly. Arguably, it was never successful in achieving any of the objectives for which organizations have paid billions of dollars. The product management approach to security is like trying to change the wheels while the car is doing a 100 mph. You won't be able to do it and you WILL get hurt along the way.

Learn more by visiting our web site: Acreto.io — On Twitter: @acretoio and if you haven’t done so, sign up for the Acreto Crypto-n-IoT podcast. You can get it from Apple – Google or your favorite podcast app.

About Acreto IoT Security
Acreto IoT Security delivers advanced security for IoT Ecosystems, from the cloud. IoTs are slated to grow to 50 Billion by 2021. Acreto’s Ecosystem security protects all Clouds, users, applications, and purpose-built IoTs that are unable to defend themselves in-the-wild. The Acreto platform offers simplicity and agility, and is guaranteed to protect IoTs for their entire 8-20 year lifespan. The company is founded and led by an experienced management team, with multiple successful cloud security innovations. Learn more by visiting Acreto IoT Security on the web at acreto.io or on Twitter @acretoio.