We all know what botnets are (think so), but anyway let’s see a proper definition of botnets taken from shadowserver… and I quote:
A botnet is a collection of computers, connected to the internet, that interact to accomplish some distributed task. Although such a collection of computers can be used for useful and constructive applications, the term botnet typically refers to such a system designed and used for illegal purposes. Such systems are composed of compromised machines that are assimilated without their owner’s knowlege.
Among the DDoS usage of botnets there are also know usages like:
Keylogging is perhaps the most threatening botnet feature to an individual’s privacy. Many bots listen for keyboard activity and report the keystrokes upstream to the bot herder. Some bots have builtin triggers to look for web visits to particular websites where passwords or bank account information is entered. This gives the herder unprecendented ability to gain access to personal information and accounts belonging to thousands of people.
Botnets can be used to steal, store, or propogate warez. Warez constitutes any illegally obtained and/or pirated software. Bots can search hard drives for software and licenses installed on a victims machine, and the herder can easily transfer it off for duplication and distribution. Furthermore, drones are used to archive copies of warez found from other sources. As a whole, a botnet has a great deal of storage capacity.
Botnets often are used as a mechanism of propogating spam. Compromised drones can forward spam emails or phish scams to many 3rd party victims. Furthermore, instant messaging accounts can be utilized to forward malicious links or advertisements to every contact in the victim’s address book. By spreading spam-related materials through a botnet, a herder can mitigate the threat of being caught as it is thousands of individual computers that are taking on the brunt of the dirty work.
and the one I’m gonna focus on (well, something derived from it) -> Click Fraud
Botnets can be used to engage in Click Fraud, where the bot software is used to visit web pages and automatically “click” on advertisement banners. Herders have been using this mechanism to steal large sums of money from online advertising firms that pay a small reward for each page visit. With a botnet of thousands of drones, each clicking only a few times, the returns can be quite large. Since the clicks are each coming from seperate machines scattered accross the globe, it looks like legitimate traffic to the untrained investigator.
My point is that many herders (botnet organizers) use a pretty raw Click Fraud mechanism, mainly just issue the command to the bot to retrieve the page and it’s advertisement and rebuild a query string to the advertisers website with the referer header set… as mentioned in the definition this may seem sometimes legitimate traffic to some, but big advertising companies would notice that something isn’t right, stuff like hundreds of clicks at (almost) the same time and similar scenario’s…
The new approach (better) would be to generate only website traffic at random hours because highly visited websites use pay-per-post campaigns (more info about pay-per-post)… and there are also other advertising systems like simple banner/ad placement on the website/blog and via the traffic stats you get paid…
How could botnets help? Well botnets would act as general users/viewers of the blog/website thus making legitimate traffic… masked by a randomized visit system… a general scenario:
- the herder issues the command to visit a website
- each bot receives the command, enters a random delay before executing it (in minutes) (ex: rand(60))
- the bot finally executes the visit and resets the delay time before revisit adding a day to it also
A very raw implementation could be easily implemented but varying from botnets to botnets, because some botnets are simple IRC based while others not…
It’s unethical… to whom?! to advertising companies only…