Don't Let a Chatbot Pick Your Ballot

Election season brings a new habit: typing "who should I vote for?" into an AI chatbot. The appeal is obvious. You get a tidy answer in seconds, without reading platforms, voting records, or local ballot measures. But that shortcut carries real risks, and it is worth understanding them before you hand over a decision that belongs to you.

Chatbots do not look up facts the way a reference book does. They generate text by predicting what a plausible answer looks like. That means they can describe a candidate's position from years ago, mix up two people with similar names, or invent a quote or a vote with complete confidence. A polished, well-organized response can hide a mistake that would be easy to catch if you checked the source.

The tools also reflect decisions made by the companies that build them, from the data used in training to the rules that govern how they handle political questions. Even when a bot aims for balance, its framing can still nudge you in one direction. And because answers shift with the wording of a question, asking the same thing twice may produce different advice. A recommendation that changes when you rephrase it is not a stable guide.

None of this means chatbots are useless for civic life. They can explain how a ballot measure would work, summarize the difference between two offices, or suggest questions to ask a candidate. Use them to get oriented, then verify what matters. Candidates' own websites, official voter guides, nonpartisan election offices, and the actual text of proposed laws are far more reliable than a machine's summary of them.

Finally, choosing a candidate depends on your priorities, your community, and what you value most. A tool can help you organize that thinking, but it cannot weigh those things for you. The decision is yours, and it is worth making with your own research rather than a chatbot's confident guess.