While the 5,000 reports provide anecdotal evidence, the AG's office performed an analysis of the 23 million public comments in order to figure out how many were submitted under falsely assumed identities.
Many comments for and against net neutrality rules are identical because advocacy groups urged people to sign form letters, so the text of a comment alone isn't enough to determine if it was submitted by a real person.
The AG's office thus examined comment text along with other factors, such as whether names matched lists of stolen identities from known data breaches. Schneiderman's office also told Ars that it looked into whether or not the submission of comments was in alphabetical order, one after another, in short time periods. In general, analysis of formatting and metadata played a role in the analysis.
The number of comments believed to be fake has grown as the A.G.'s investigation continues, and it isn't done yet. Schneiderman's office is still analyzing the public comments. We asked Schneiderman's office how many of the fake comments supported net neutrality rules, and how many opposed them, but were told that the information was not available.
While fake comments used names and addresses of people from across the nation, more than "100,000 comments per state" came "from New York, Florida, Texas, and California," Schneiderman's announcement said.