Social Security Numbers are Predictable, Study Finds

A new study out of Carnegie Mellon University finds that Social Security Numbers can be easily predicted. The first five digits of a SSN are based on the state where a person is born and the date of birth. In less than 1,000 attempts researchers at CMU used publicly available information to determine all nine SSN digits for 8.5 percent of individuals who died between 1988 and 2003.

Dr. Alessandro Acquisti says they expected to see patterns, but were surprised by how easy it was to predict the numbers. He says the study provides conclusive evidence that the numbers should not be used as identifiers and passwords. He says they are not the first to call on the Social Security Administration to find a new way of identifying individuals. He says randomizing the numbers could be a short-term solution, but should find another system to replace all SSNs in the future.