Red team testing: hacking and lying your way to data

Scanning for vulnerabilities

We discovered some potential digital vulnerabilities as well. Firstly, their website was running a vulnerable version of WordPress, and although we couldn’t fully compromise the server as it had been well hardened, we were able to write new pages and sites for it.

Based on this, we decided on a two-pronged attack:

1. We’d create a phishing portal on the hacked WordPress server designed to look like their Outlook Web Access page.
2. One of us would visit the physical offices with a photoshopped ID badge based on the one we found on Twitter, wander onto the premises and try and attach malicious hardware to the network, or to any machine we could get close to.

Setting up remote access

Our phishing portal was very hard to spot as it was well disguised on its own server, on the company domain and complete with a valid SSL certificate. Several workers were caught out by it, and this gave us access to their accounts. The client didn’t expose any remote desktop or cloud-based services for us to use, so we hung onto these stolen credentials for later…

Gaining entry

Thanks to modern smartphones being so advanced, the images of employee badges were good enough for us to replicate their appearance. While they looked legitimate enough at first glance, we didn’t know what technology we needed to access the automatic door system. We created several kinds of cards based on known electronic badge types. These didn’t work, but one using the same technology made a warning ‘beep’ sound, which we used to add a sense of authenticity.

Once I arrived on the premises, a quick look around the general area confirmed that there was indeed a back door. However, it was a fire exit adorned with a sign that read ‘this door is alarmed’. As I was trying to be inconspicuous as possible, I decided to go through the front, non-alarmed door. This was obviously a bold move as it meant marching right into the line of sight of a security guard, who wouldn’t recognise me and could ask questions I was completely unprepared for.

Now, this truly separates red team testers from penetration testers. I strolled in, carrying several bottles of milk I purchased previously. I nodded to the security guard and gestured to the milk in my hands, saying ‘milk run’. Next, I swiped my card, triggering warning beeps. I repeated the process, pretending to get annoyed with it not working until the security guard came over.

“This thing hasn’t worked properly since I put it through the wash”, I said. With a pleasant chuckle, the security guard swiped me through and even pressed the call button on the lift for me. From here, blagging my way into the office itself was easy, and I even dropped off the milk in the kitchen area. After all, if I’m about to steal a load of information, it’s the least I could do!

The exploitation phase

Once in the office, I took out my laptop (complete with large Wi-Fi antennas) and strolled around with an intense look of concentration. Fortunately, if you work in IT, most people are reluctant to speak to you unless they absolutely must, so I was able to make my way to a partially empty office and plug a disguised USB ethernet adaptor into an unmanned desk PC.

Our USB ethernet adaptor contained a small Linux machine that would remotely connect to our own servers and allow us to tunnel into the target network. However, it seemed the internal network had some form of filtering in place that blocked unknown devices, which would have meant that the entire endeavor could have ended there, (wasting several pints of perfectly good milk). You’d be forgiven for thinking this meant game over – but I wasn’t giving up easily. Remember, we were also armed with several user credentials we’d obtained via our phishing portal. After logging into the computer, we were able to bypass this filtering and gained access to their servers.

From here on in, the rest is mostly the same as any internal penetration test. We found things that were broken, abused them, stole passwords, spread further and kept on going and gathering evidence until there was nowhere else to go or until we were caught. We weren’t caught on this occasion. But as part of this experiment, we provided our client with a comprehensive report and helped them to tighten their processes and improve their cybersecurity. We demonstrated how far a malicious threat could be willing to go to gain access to their business. It’s not always someone at a laptop on the other side of the world. Sometimes it’s simply a man with some milk.