How Algorithms Can Help To Spot Voter Fraud

The last few years have seen a marked increase in concern around the integrity of elections, not least due to accusations that Russian hackers tried to access voter records during the last presidential election.  This has been joined by concern that administrative errors can further distort electoral reality.

New research from Caltech aims to deploy AI to track voter data and better identify rogue patterns.  The algorithm was trained on over 1.5 million voting records from Orange County, and the team believe their system can accurately assess the integrity of any election.

“Manipulations to voter records can wreak havoc on an election,” the researchers say. “You could have people showing up to vote who are not on the list, or people’s addresses can be changed in the databases so that voters do not get their instructions in the mail. There are many scenarios, some fraudulent and some administrative, that can negatively influence the quality and integrity of elections.”

Electoral integrity

The algorithms were designed in such a way as to take snapshots of voting records on a daily basis.  This is important as voting records are fluid as voters are added and removed from the system, addresses are changed and so on.

The team built a couple of algorithms, one of which was designed to measure these dynamic changes to the voting record, and then a second to try and spot any statistical anomalies in this process.

A third algorithm was then build to look for duplicates in the voting records.  These don’t have to be the result of nefarious actions, and are often a common administrative occurrence as voters move or register in multiple places.  It nonetheless provides a marker for the overall health of the database.

The team hope that not only will their work be valuable in itself, but it will also be a launchpad via which other projects can take place to enhance the integrity of voting processes.

“Administrative errors and the potential for shenanigans loom large in U.S. elections, and we need to fend off possible corruptions to voter rolls in a timely manner,” they say. “Our vision is to have all states upload voter data on a daily basis and to have algorithms monitor their integrity.”

Facebooktwitterredditpinterestlinkedinmail