All those pet cat profiles add up: Facebook has reported that roughly 76 million of the 1.06 billion user accounts on its social network are bogus in some way or other.
The company identifies three types of accounts that don’t represent actual users: duplicate accounts, misclassified accounts and undesirable accounts. Together, they added up to just over 7 percent of its worldwide monthly active users last year.
Facebook disclosed the figures in its annual report filed with the U.S. Securities & Exchange Commission on Friday.
Duplicate accounts, or those maintained by people in addition to their principal account, represent 53 million accounts, or 5 percent of the total, Facebook said.
Misclassified accounts, including those created for non-human entities such as pets or organizations, which instead should have Facebook Pages, accounted for almost 14 million accounts, or 1.3 percent of the total.
And undesirable accounts, such as those created by spammers, rounded out the tally with 9.5 million accounts, or 0.9 percent of users.
Facebook said it continually tries to improve its ability to identify these duplicate or false accounts. It also noted that there’s a higher percentage of such accounts in developing countries such as Indonesia and Turkey, compared to developed markets like the U.S. and Australia.
Using a fake name is against Facebook’s policies and it encourages users to report friends who use false names or set up fake accounts.
“We have a dedicated User Operations team that reviews these reports and takes action as necessary,” along with technical systems in place to flag and block potential fake accounts based on name and anomalous site activity, a company spokeswoman said.
The latest figures aren’t very different from those Facebook reported midway through last year. There were slightly fewer duplicate accounts for the quarter ended June 30. But there were almost twice as many misclassified accounts, at 22 million, and also more undesirable accounts, at 14 million.
Facebook said the numbers it reports are based on a limited sample of accounts, and that it exercises judgment in its counting, “such as identifying names that appear to be fake or other behavior that appears inauthentic to the reviewers.”
Still, Facebook does not specify in the filing how it judges the authenticity of one “meow” over another.