The crime rate began to grow exponentially, however, as actors, athletes, musicians, politicians, and other big names opened their Twitter accounts during November 2008 through April 2009. By that April, the number of malicious Twitter accounts grew 66 % to 3.36%. By October 2009, by Barracuda Lab's estimates, the number of malicious accounts reached 12%. That means about one in eight accounts created were found to be malicious.
Some of the more notable attacks throughout last year included Guy Kawasaki's account offering a celebrity sex tape, Koobface malware activity increasing last summer, as well as distributed denial of service attacks and DNS records compromised by the "Iranian Cyber Army" in December.
It's interesting to see it quantified how, as Twitter grew mainstream, the attackers followed.
A big part of the security challenge on Twitter are phishing accounts setup by fraudsters who rely on users' clicking on shortened URLs that obfuscate the actual destination. Another problem is legitimate accounts that get hijacked and spew direct message phishing attacks and spam.
To combat those threats, Twitter announced on its blog yesterday that it will direct shortened links sent in direct messages and e-mail notifications through a new back-end service that promises to detect, intercept, and stop the spread of malicious links.
That's a good first step, but expect the attackers (just as they do with malware and spam) to quickly adjust their tactics to bypass the new filter.
The Barracuda Labs 2009 Annual Report, available for download as a .PDF, also shows statistics on how e-mail-based malware grew by threefold in the second half of 2009, and that spam focusing on jewelry grew 10 times during 2009 - as the price of gold rose. The report also found that Web exploit kits, rogue ant-virus software, and search result poisoning were the most popular forms of Web-based attacks.