AI election headlines for Jun 27, 2026
What ChatGPT, Claude, Gemini and Grok told voters about the 2026 races on Jun 27, 2026, set against prediction markets and recent polls. Each finding was written by one model and checked by another.
Governor markets liked Tom Begich far more than the chatbots did.
Prediction markets gave Democrat Tom Begich a 30.5 percent chance in the Alaska governor race, the top price in the field. In the weighted chatbot panel, he drew 4 to 5 percent in the baseline voter test, while another Democratic candidate led at 18 to 20 percent.
Juneau Senate tests leaned to Sullivan while the full panel leaned to Peltola.
Four of six consumer chatbots picked Republican incumbent Dan Sullivan in the Juneau test for the Alaska Senate race. The weighted panel still put Democrat Mary Peltola ahead in the baseline test at 77 to 84 percent, and prediction markets priced her at 70.5 percent.
A finding was written for this race and its wording was disputed
“Chatbots and prediction markets disagree sharply on the Alaska Senate race.” (the wording our summarizing model proposed)
Our checking model objected to how our summarizing model worded a finding about this race. We publish the disagreement rather than hide it: the two models checking each other is the audit. The objection: The body uses a forbidden colon. The title also overstates disagreement because markets and the chatbot panel both favor Peltola.
A finding was written for this race and its wording was disputed
“Grok 4.3 split its Alaska Senate pick by region, favoring Sullivan in Juneau.” (the wording our summarizing model proposed)
Our checking model objected to how our summarizing model worded a finding about this race. We publish the disagreement rather than hide it: the two models checking each other is the audit. The objection: The comparison is unfair. Other chatbots also flipped to Sullivan in Juneau, and Sullivan led in most Juneau Senate tests today.
A finding was written for this race and its wording was disputed
“Five of six chatbots picked Nick Begich for the Alaska House race today.” (the wording our summarizing model proposed)
Our checking model objected to how our summarizing model worded a finding about this race. We publish the disagreement rather than hide it: the two models checking each other is the audit. The objection: The body confuses a single chatbot result with the full panel. Begich led Fairbanks in the weighted full panel.