Back 
Netskope Threat Research Labs recently found an Amazon-themed phishing page hosted by Azure App Service. The phishing page was created with an intent to steal Amazon credentials and other sensitive personally identifiable information (PII) from the victims. In an earlier post, phishing in the public cloud, we detailed phishing campaigns that used public cloud hosting services to serve up parts of the attack. Here, the entirety of the attack is hosted solely in Azure. This post describes how the phishing page was created, highlighting how easily attackers can use public cloud services to create convincing phishing pages. It also outlines how the attack looks to the victim and provides some more details about the attacker.
Hosting Phishing Pages on Azure App Service
Azure App Service is a platform that enables users to easily deploy websites. The attacker here used Azure App Service both to create the phishing web form used to collect data from the victims and to create the database used to store the data. The following are a few of the reasons that the attacker might have chosen Azure App Service.
- The low cost of hosting the phishing site. A new Azure subscription currently includes offerings like 12 months of free services and a $200 credit. These enable an attacker to host a phishing page at no cost to them.
- Microsoft SSL certificates and domains. A site hosted in Azure App Service will be hosted from a Microsoft domain and have a Microsoft-issued SSL certificate. Both of these tend not to arouse suspicion in victims. Furthermore, the Microsoft certificates and domains tend to help attackers evade traditional security solutions as well.
- Accessibility. All that is required to sign up for Azure is a valid credit card number; this makes it easy to create an account and setup a phishing page.
- Ease of deployment. If a site gets shut down, it is very easy to spin it up again in a new account.
While these points were written specifically about Azure, they also apply to the other major cloud service providers, such as AWS, GCP, and Alibaba. We expect that all of these services will be used by attackers to create similar phishing sites.
Netskope Detection
Netskope for Web detects and blocks the phishing page as shown in Figure 1.
Disclosure
The phishing site we discovered was reported to Microsoft on 14 June 2019.
Amazon themed Phish Analysis
On visiting the phishing page, the victim is presented with a message that their Amazon account is temporarily suspended and prompted to validate their account by re-entering personal information, as seen in Figure 2.
Upon entering the details, the credentials are sent to cr.php page hosted in emazon.azurewebsites[.]net as shown in Figure 3.
The packet capture illustrating this action is shown in Figure 4.
The attacker presents messages to the victim that the connection is established to Amazon and that verification is successful. The victims are then redirected to amazon.com. This process is depicted in Figure 5.
Attack Kill chain
The visual depiction of attack kill chain of the phishing attack is shown in Figure 6, beginning with the victim visiting the phishing page, entering their information, and being displayed multiple messages before being redirected to amazon.com. The entire kill chain is hosted in Azure App Service until the final redirection.
Threat Actor Information
The phishing page we discovered has been active since the first week of June. We were able to identify some contact information for the author of the phishing page, but were unable to tie that contact information to any other phishing campaigns. We will continue to monitor this actor to see if we can tie them to any past or future phishing campaigns.
Conclusion
This post described how an attacker used Azure App Service to create a convincingly realistic phishing page. We expect they chose Azure App Service for its low cost, reputable domain and SSL certificate, accessibility, and ease of deployment. We also expect to continue seeing these campaigns launched from all the major cloud providers.
To avoid becoming a victim of a campaign like this, there are a few red flags to look out for. First, websites like Amazon generally do not ask for the detailed information the attacker was requesting here. Second, the domain is one used by Azure App Service, not the amazon.com domain where a user would normally enter their credentials.
Read the blog