The reality of how much malicious content is on Twitter was demonstrated at the RSA Conference in San Francisco.

In a presentation called 'The Dark Side: measuring and analysing malicious activity on Twitter', Barracuda Networks' research scientist Daniel Peck and chief research officer and VP of cloud services Dr Paul Judge claimed that with any website with user generated content there is a gap in trust among users, some of who are genuine and others who use them as a means to attack other users' accounts.

Judge said: “Twitter is so open that it allows attackers to make use of it. With account hijacking, what risk model is there? There are security flaws and malware in the content and you can see hijacking or simply spam.”

He pointed to incidents such as where compromises forced users to follow another user, the 'onmouseover' cross-site scripting attack where users tweeted a phrase by simply clicking on a piece of code in a tweet and various account hijacking incidents.

Peck said that millions of users accounts are created every day and for every 100 users, only one per cent had 1,000 or more followers. “The site saw huge growth in 2009 and half of the most popular users joined in this period, as did those who like to follow them. The crime rate also increased by 66 per cent and there was a large spike with the FIFA World Cup, with everything from spam to malicious stream viewing being promoted,” he said.