Cybercriminals have begun abusing legitimate cloud services in new ways to try and sneak attacks past security controls and make their scams appear more convincing to intended victims.
Netskope on Monday said its researchers had observed a recent trend among attackers to send phishing emails and SMS messages with links to malicious sites and content hosted on cloud services such as Amazon Web Services (AWS), Microsoft Azure, Alibaba Cloud, and Google Docs.
The security vendor says it has seen the technique being used to try and direct users to scam pharmacy sites, dating sites, and tech support sites, designed to steal personal information or for blackmailing victims. In other instances, attackers are abusing Google Docs to create and share presentations that contain malicious links.
The scams themselves are old and well-known. What's new is the use of cloud services to host them.
For cybercriminals, cloud infrastructure services such as AWS and Azure provide a cheap, dynamic hosting option, says Abhinav Singh, threat researcher at Netskope. These services provide all of the native features of the Web, including Web hosting.
"Traditionally, these scams were hosted by registering new domain names or by compromising existing ones," Singh says. "With cloud’s static object stores, you get access to creating an infinite number of domain names with just a few clicks."
An object store on a cloud infrastructure-as-a-service platform (IaaS) is a storage system for storing and delivering content to users. Unlike the case with block storage, content stored in object storage can be accessed via the Web or through APIs.
"When you create an object store, you give it a name, and users access the objects using that name," Singh says. An AWS S3 storage bucket that is named "my-text-bucket," for instance, would be accessible via my-test-bucket dot s3 dot Amazonaws dotcom, he says.
Hosting a malicious site or content in a cloud object store eliminates the need for attackers to have to keep registering domains as existing ones get discovered and blocked. Attackers can simply use a domain generation algorithm and create a new object store and URL as needed and on a continuous basis, Singh said.
Because these malicious sites are hosted on trusted cloud service infrastructure, any links to them appear more genuine to users and are able to bypass content filters at the same time. Exacerbating the situation is the fact that many cloud service providers appear unable to identify and take down fake and malicious sites quickly enough. "Based on our analysis of these scams, we have seen them to be alive and running even weeks after they went live," Singh says.
According to Netskope, attackers are currently using cloud services largely for hosting well-known and long-running scams targeting individuals. But the same techniques could easily be used to target enterprises that use services such as Google Drive, the security vendor said.
"We believe that the trend of abusing cloud services by attackers is only going to increase in the future," Singh predicts. The use of cloud assets for cryptomining, malware hosting, and phishing is already common, he says.
Organizations need to start educating users about the dangers of blindly clicking on links to cloud services. They also need to put filters and policies in place for detecting and blocking cloud-based attacks, he says. Based on the change in tactics, "it is clear that attackers are trying to adapt to the changing landscape of the Web in order to victimize as many end users as possible," Singh said.
Join Dark Reading LIVE for two cybersecurity summits at Interop 2019. Learn from the industry's most knowledgeable IT security experts. Check out the Interop agenda here.