Cambridge Analytica has modeled individual peoples' personality and uses AI to influence their voting behavior based on individual personalities.
Any company can aggregate and purchase big data, but Cambridge Analytica has developed a model to translate that data into a personality profile used to predict, then ultimately change your behavior. That model itself was developed by paying a Cambridge psychology professor to copy the groundbreaking original research of his colleague through questionable methods that violated Amazon’s Terms of Service.
By correlating subjects’ Facebook Likes with their OCEAN scores — a standard-bearing personality questionnaire used by psychologists — the team was able to identify an individual’s gender, sexuality, political beliefs, and personality traits based only on what they had liked on Facebook.
After a successful proof of concept and backed by wealthy conservative investors, Analytica went on a data shopping spree for the ages, snapping up data about your shopping habits, land ownership, where you attend church, what stores you visit, what magazines you subscribe to — all of which is for sale from a range of data brokers and third party organizations selling information about you. Analytica aggregated this data with voter roles, publicly available online data — including Facebook likes — and put it all into its predictive personality model.
Nix likes to boast that Analytica’s personality model has allowed it to create a personality profile for every adult in the U.S. — 220 million of them, each with up to 5,000 data points. And those profiles are being continually updated and improved the more data you spew out online.
Point being, with these methods, it should be possible to accurately estimate each person's likelihood of voting and who they would vote for. You could do some fuzzy voter roll matching to also determine if they are registered to vote, or you could use their personality and engagement to estimate whether they are registered to vote. If you combine that with strategic political information like past voter turnout, electoral college facts, polling data, etc.; then you should be able to accurately predict election outcomes better than conventional polling does.
Although, to my knowledge, Cambridge Analytica wasn't themselves making an effort to predict an election outcome. Predictions notwithstanding, they were trying to model a voter's personality, and then use bots and targeted ads to present them with messages that were strategically aimed at being persuasive towards their personality model.
Simply using historical turnout models can't possibly be accurate, because it ignores demographic shifts and turnout shifts. For example, Democrats turned-out for the 2018 mid-terms much more than they did for past mid-terms, because they were much more galvanized. You can't predict that sort of year-to-year passion using only historical models. The best way to attempt to is to use voter rolls to gauge first-time voter turnout, and if you're lucky, the amount of first-time voters can also give you an accurate idea of the turnout of repeat voters.