Hidden risks of advertisements

Wednesday 30 September 12:00 - 12:30, Red room

Doina Cosovan (Security Scorecard)
Catalin Lita (Security Scorecard)

Adware is everywhere. Mobile applications are widely used and most of them, especially the free ones, embed adware software development kits (SDKs). Web browsing frequently exposes users to adware, as well.

There are multiple methods for sinkholing adware infrastructure. First, adware developers might let their hard-coded domains expire. Second, they can use wrong domains, intentionally or unintentionally. Third, some adware SDKs have started to use domain generation algorithms (DGAs) as a fallback mechanism for the hard-coded domains. This seems an attempt to bypass ad blockers, which blacklisted their hard-coded domains. In the last case, using an ad blocker turns out to be more dangerous as it exposes the user to sinkholable domains, which might be under an attacker’s control.

This presentation / paper focuses on analysing the security risks involved when adware infrastructure can be sinkholed. First, an attacker can passively gather, for later use, personally identifiable information about both the users and the advertisers. Second, an attacker can actively serve specially crafted advertisements, including malvertising, in order to exploit and infect the contacting systems.

We sinkholed multiple adware-related domains, but we are going to illustrate a few interesting use cases.

The first use case is an Android mobile adware with an expired hard-coded domain, contacted by more than half a million unique IP addresses daily from tens of different applications.

The second use case is a multi-platform mobile adware mediator with a fallback DGA, which is used in more than a thousand different applications and receives requests from tens of thousands of unique IP addresses (note that this is only the subset of IP addresses that didn’t manage to contact the hard-coded domain out of the total number of IP addresses using the adware mediator).

The third use case is a browser HTML5 adware player which intentionally used a non-resolving domain in order to evade syntax errors. They fixed the issue a few days after we sinkholed the domain, but during those few days, the sinkhole received requests from half a million unique IP addresses.

The fourth use case is an adware platform for browsers, whose domain was inserted in the SDK by the developers with a typo. It is contacted by more than two million unique IP addresses daily and is used by more than ten thousand different websites.

For one of the presented use cases, we created a proof of concept illustrating how an attacker can serve a malvertisement by owning an infrastructure domain. The most concerning aspects are: this happens regularly, this happens for the latest official adware platforms / SDKs, and this happens for adware that have a large user base.

 

Doina Cosovan Doina Cosovan has a computer science degree. From her second year of college she worked for Bitdefender's malware research team, before joining Security Scorecard five years ago. She has presented at conferences such as Virus Bulletin, Caro and AVAR. Some of her interests include malware, botnets, reverse engineering and machine learning.

Cătălin Valeriu Liță Cătălin Valeriu Liță received a Bachelor's degree in computer science from the Technical University Gheorghe Asachi, Romania, Iasi, Faculty of Automatics and Computer Science. He has a Master's degree in information security from the Alexandru Ioan Cuza University of Iași, Faculty of Computer Science, a Master's degree in business administration from the Alexandru Ioan Cuza University of Iași, Faculty of Economics and Business Administration, and a Ph.D. in computer science from the Faculty of Computer Science. He has presented at CARO and Virus Bulletin conferences. Prior to joining Security Scorecard he worked for nine years in Bitdefender's anti-malware team.

Back to VB2020 Programme page

Other VB2020 papers