Facebook's disclosure that it found 14 million user accounts it considers "undesirable" - meaning they are likely spewing spam or other malicious links and content - highlights the need for better user verification on the popular social network, security experts said.
In a US Securities and Exchange Commission (SEC) filing, Facebook said this week that as of June 30, there were 83 million fake accounts on the site, or 8.7 percent of the total of 955 million. Of those bogus accounts, 1.5 percent, or more than 14 million, were classified as "undesirable" and set up "for purposes that violate our terms of service, such as spamming."
The disclosure highlighted once again the security risks on the site for users. Spam remains a problem, as well as malicious links that download malware onto users' computers. "Fourteen million is definitely a high number," Xuxian Jiang, a security researcher at North Carolina State University, said in an interview via e-mail.
Facebook did not immediately respond to a request for comment.
Security experts agree that Facebook will need better user-verification tools to help combat the problems associated with bogus accounts. "One solution off the top of my head would be enforcing a real-identity rule for social networks, if not, at least some sort of user verification," Jiang said. "But on the other end, my uneasy concern is that it will sacrifice online privacy to some extent. A delicate balance may need to be in place."
Atif Mushtaq, a security researcher and botnet expert at FireEye, said Facebook and other websites are battling the ineffectiveness of the "captcha" test used to ensure a registrant is a real person and not the computer of a spammer or hacker. The challenge-response test typically asks for the registrant to type in a random series of letters and numbers.
Cybercriminals are getting around this test by using computers they have already hijacked. When they need a real person to type in a captcha request, they send a bogus pop-up window asking one of their victims to input the needed letters and numbers to prevent their Windows operating system from restarting, Mushtaq said.
That type of captcha workaround has become popular with botnet operators. "There's no short-term solution for this problem," he said.
Facebook has taken a number of steps to try to reduce malicious links and malware on the site. The company continuously monitors the site for suspicious content and will notify users if it notices unusual activity from their computers. The site also offers an Anti-Virus Marketplace where users can download free AV software from security vendors.
Of the remaining fake accounts, Facebook reported that 45 million were duplicate accounts and 23 million were categorized as "user-misclassified," which means they were personal profiles set up for businesses, boats, pets or some other non-human entities. The disclosure comes as Facebook is under increasing scrutiny by advertisers, some of who have complained that Facebook's ad model is ineffective.
From an enterprise perspective, the presence of so many spam- and malware-distributing accounts offers further justification for corporate controls on the use of Facebook, as well as the use of security tools, such as malware detection and intruder prevention and detection technology.