In the past we have talked about the internal assessments that we perform here at Digital Shadows. As part of one of our regular tests back in November 2018 the team acquired a domain for a major online service: The only difference from their primary domain was the Top Level Domain (TLD). For example, instead of the common TLD “.com”, we could purchase something like “.io” and use this for activities like internal phishing exercises. The best thing about it is you can pick one up for under £15 per year. What a bargain!
Since we are slowly approaching the one-year anniversary of our proud purchase, we wanted to talk about how easy it is to conduct domain squatting and typosquatting, and how little monitoring still goes on around them in the industry.
Let’s take a moment to discuss some of the differences between the two techniques commonly used: Typo and domain squatting.
What is a Typosquat Domain?
We all know what a typo is; and if you’re honest and not blaming your phone’s autocorrect, we make them a lot. The main use of typos when it comes to phishing is that when we create a variation of a reputable or trusted domain using a deliberate typo in the address, we can switch out an “i“ for a “l”, or add or remove characters that would be easily missed. Quite simply, the technique capitalizes on users hitting the wrong key when inputting a URL. Like outlined in our blog on Facebook’s cryptocurrency Libra, attackers can also register domains with characters from other alphabets which resemble letters in the Roman alphabet.
Figure 1: Example of typosquat domain
There is a huge amount of variations that could be used to create a convincing looking domain. While some of these may seem obvious and suspicious, try spotting the difference in your Outlook client or browser on your mobile device. They can be very sneaky indeed.
What is a Domain Squat?
A domain squat requires less creativity but can be even more deceptive. With a domain squat we are looking for a domain that directly represents an organization’s brand or service. We have all seen scam emails where an attacker used a domain that contains a brand or company name that we know and trust. For example, domain squats of services like PayPal are common: You could easily acquire a domain such as “paypal-fraudteam.com” and move on to creating your lure and getting ready and standing up your infrastructure.
A domain squat can take many forms, but it will usually include a brand, service, or company name that closely resembles that of the legitimate one.
Figure 2: Example of domain squats (Source: Digital Shadows SearchLight)
Between these two techniques, an attacker has many options. All you need is some creativity, some tech skills, or the willingness to do a little bit of copy and pasting from various guides floating around the Internet.
Figure 3 – Be creative and plan your lure
Phishing Campaigns: Landing your fresh phish
Often with phishing campaigns we use landing pages, which are typically configured using domain names that mimic the brand or company of a reputable organization. The purpose of a landing page may vary: Most commonly though it is used to deliver a payload of some form or to host a spoofed login page. Using the right domain name, whether that is a domain squat, typosquat, or another technique, you can create a masterpiece that can test the security awareness of even your most tech-savvy team members.
Landing pages are just one part of it though. Being able to use the domain as the sender for emails will also add a lot of credibility to the phishing email. In previous blogging, we have talked about email spoofing and methods of reducing the risk of these types of attacks.
Spoofing has more associated risks these days than it once did. We could use a random free email provider, but when you are testing security awareness against a more “sophisticated” attacker, you want to pull out all the stops to have maximum chances of a successful phish.
Figure 4 – Spoofed login page
How Attackers Use Spoof Domains
So what can we do with this domain now that we have spent our hard earned £15 on it?
Our internal goals in this scenario was to gauge and help improve overall user education and security awareness. With a convincing domain in hand, we thought we may only have a short window to set up the assessment landing page spoofing the actual service, create the lure, set up the email services and phishing platform, and then run the campaign before receiving a takedown notice or email from service providers.
However, our experience in typo and domain squat detection may have been overly optimistic. Our internal assessment ran over two days with some great results and feedback all around.
This experiment shows that with a little bit of creativity, effort, and a good internal security team, you can run a smooth campaign with minimal setup. In this scenario we were using an in-house setup comprising of:
- Our newly acquired domain – DNS configure
- Self-hosted Ubuntu Linux server
- Postfix – Mail server
- GoPhish – Open Source Phishing Framework
- Plenty of patience
Even if you do not have a large team or budget, small setups like this are achievable for most organizations.
Let’s not encrypt
While setting up the infrastructure, one of the decisions was whether we should set up a secure connection on the phishing landing page. While this is more realistic when simulating a more technically capable attacker, there are also some issues around doing so.
Websites using secure connections will need a certificate to be issued by a trusted third-party known as a Certificate Authority (CA). With most CA’s now, when a certificate is issued, a certificate transparency log entry for the creation and issuing of the certificate is created. This is part of an open framework set up to allow for the auditing of certificate creation. Due to this, anyone can monitor for a certificate being created. By looking for keywords such as brands or trademarks, it is possible to identify the existence of spoofed domains.
Figure 5: Certstream
The internal struggle was real: If we use HTTPS for the landing page, we increase the chances of successfully phishing our targets (people are inherently more trusting of a site which uses HTTPS as opposed to one with HTTP), however by generating a certificate we could also potentially setting off red flashing lights and sirens in a Security Operations Centre somewhere.
So we decided, let’s not encrypt this time.
How to Detect Typosquats and Domain Squats
So how can we detect when someone has registered a domain or typosquat? There are a few ways this could work; some of the most successful ways are as follows:
- Generating variations of your domains or trademarks/brands and checking for registration and DNS records
- Monitoring phishing reports for domains spoofing your trademarks or brands
- Monitoring certificate transparency logs for the creation of certificates, looking for your trademarks or brands
Figure 6: Our SearchLight platform offers domain monitoring – seen here
Digital Shadows offers domain monitoring like what we’ve described here, for any company domains that you (as a client) would want to keep an eye on. Once SearchLight detects a potential squat for a client’s domain, a risk score will be assigned to an alert that is sent to the client. The risk score varies depending on how likely a domain is to be used within a targeted attempt against the company (such as a phishing email).
Additionally, our monitoring capability does just that: monitors that domain and changes that it may go through over time. Say for instance, the domain did not originally have an MX record assigned to it, but after a month one was added. This is a sign that emails can be sent using the domain and more likely that it could be used for phishing.
Internal security exercises are not only useful for raising the security awareness of our employees, but also for looking at the offensive operational issues around what constraints we may have as a red team. For our team, we look at each assessment we do through a purple lens. While the assessments are being carried out, we are also looking at our logs and services. If something got through, how could we detect it next time? How can we make the process easier for us to detect and report incidents in the future?
Simple exercises such as the one described in this blog are not out of reach even for the smallest of teams.
Looking at the issue in hand (domain and typosquats), simple implementations looking at services such as CertStream can give you some good visibility over what domains are having certificate issues that may be abusing your trademark or brand. Organizations can also generate variations of their domains and monitor them for changes and the addition of suspicious or malicious content. The detection and takedowns of malicious domains certainly have their challenges but is by no means impossible.
To learn more practical tips around monitoring for domain squats and typosquats, and other business exposure, such as exposed credentials, documents, and infrastructure weaknesses, check out our Practical Guide to Reducing Digital Risk.