When a voter, donor, or journalist asks an answer engine about a candidate, the answer they get is now part of the campaign. These papers make the case for treating that answer as a channel worth measuring and managing, the same way campaigns already treat television, mail, and the doorstep.
They are written by Ed Forman, founder of Raise Presence, which built Kyanos to measure how answer engines describe candidates, officeholders and advocacy organizations. We publish them because the field is new, the stakes are concrete, and most of what has been written about optimizing for AI was measured on commercial content rather than on political content.
Every factual claim carries a footnote that clicks through to its source. Where a source sits behind a paywall we pair it with one you can read.
We do not assert how these systems work inside. Their retrieval and weighting are not publicly documented, and a confident mechanism is the easiest thing to get wrong.
Measurements carry the population, the engine and the date they were taken. A number about engine behaviour is a reading on a date, not a standing fact, and we treat findings drawn from general commercial content as evidence about that content until someone measures politics directly.
A hypothesis is labelled as a hypothesis. Where we cannot verify something, we leave it out or say plainly that we could not confirm it.
Seven papers and a companion, building in sequence. They start with the scale of AI's reach into the electorate, connect AI presence to the mechanics of persuasion, apply it to the organizations that endorse candidates, then turn to the economics: where AI answers fit in a campaign budget, what that return looks like, and where remediation stops working. Each paper assumes the one before it.