We all knew there were plenty of bots lurking around online causing mischief, but new research has shed light on just how many ‘fake’ internet users there are – apparently, as much as 10% of all web traffic may in fact be driven by bots capable of mimicking humans by posting comments and even entering contests online.

Solve Media, an online advertising firm based in the US, said that its latest study was performed over the last 18 months, based on the authentications of more than 2 billion internet users on over 5,000 different websites.

Perhaps even more startling than the sheer number of bots is the rise of so-called “aberrant traffic”, which involves bots getting up to no good, performing actions like registering for websites, casting votes and entering internet contests. According to Solve Media, such traffic has risen by almost 400% since their study began.

For the average internet user, the increase in bot traffic isn’t going to make much of a difference, but for advertisers, it could prove to be fairly expensive. Solve Media say that advertisers could be spending as much as 10% of their budgets on these mischievous robots. In the meantime, publishers face the burden of having to pay for software to purge the bots or manually sort out the human comments/registrations from the fakes.

In order to deal with the bots and prevent them from wasting money needlessly, publishers are encouraged to take steps such as forcing users to complete ‘captchas’ when signing up for anything or making comments.

Ari Jacoby, Solve Media’s CEO, says that publishers have a responsibility to make life difficult for the bots:

“Once bots begin to attack a sight, they’re everywhere on the site … But once they meet significant resistance, they’re trained to go away.”