Artificial Intelligence (AI) offers new avenues to combat voter fraud. These past mid-term elections saw a historic voter turnout and probably a historic amount of effort in terms of processing all of those ballots. There is no doubt that making voting easier and more accessible is a major factor in increasing voter turnout. To encourage voting, states are increasingly turning to mail-in balloting to add to the already present absentee and military ballots.
However, when processes are made easier for customers or constituents, they also can inadvertently become more complex for those who must process the ballots. A recent article and a segment at NPR highlighted the challenge when it comes to making voting easier: the requirement to verify voter signatures.
Mail-in Ballot Processing and Voter Signatures
Absentee, vote-by-mail, and military ballots all require a signature to be matched to a verified authentic signature in order to reduce the potential for fraud when it is not possible to make an in-person verification of a photo ID. As you might expect, this process can be time-consuming and error prone for a number of reasons. The article cites problems with how signatures change over time that result in ballots being invalidated and goes on to explain that often the official signatures kept on file can be fairly old and not always representative of how a current signature looks.
Younger voters are four times more likely to have ballots rejected due to changes in the way their signatures look. For instance, my signature at age 18 is remarkably different from how I sign my signature today. Added to this problem is the fact that many election officials are not sufficiently trained on how to compare signatures. This is where automated signature verification powered by artificial intelligence comes in.
AI’s New Role
Software powered by AI such as that developed by Parascript can evaluate two signatures that look completely different and determine whether or not the signatures belong to the same person. It does this—not by looking at the signature in its entirety—but by examining portions of the signature based on a number of attributes.
In this way, the software can even authenticate a partial signature such as one that is mistakenly cropped. The software also is able to examine a candidate signature against a broad range of references in a very quick amount of time, which enables a much higher degree of precision, but without the typical extra time that an election official would need.
The article cites Colorado (Parascript’s home) as one state that has widely adopted this technology in order to support its vote-by-mail program. As more and more states work to make voting easier, automated signature verification will be a key enabler.
Vote-by-Mail | Petition Automation
Parascript software supports vote-by-mail by processing ballots and verifying signature authenticity. The software locates and extracts the ballot data, identifies if a signature is present and then provides signature verification with unprecedented accuracy. It enables signature authentication for voting by mail where it is crucial to detect signature fraud. The software provides the highest degree of accuracy for any document type that requires signature validation. Find out more here.