Skip to main content
Voting Monitor2026 Senate toss-ups and Alaska races
← All daily findings

AI election headlines for Jun 3, 2026

What ChatGPT, Claude, Gemini and Grok told voters about the 2026 races on Jun 3, 2026, set against prediction markets and recent polls. Each finding was written by one model and checked by another.

  1. Alaska U.S. House At-Large Race 2026

    Gemini 2.5 Flash flips its Alaska House pick to Bill Hill for Fairbanks voters.

    Gemini 2.5 Flash picked Independent Bill Hill at 52 percent for a Fairbanks voter on the Alaska House race today. The same chatbot picked Republican Nick Begich for an Anchorage voter at 48 percent. No other chatbot in the panel switched its top pick by region on this race.

  2. Alaska U.S. Senate Race 2026

    Gemini 2.5 Flash swung from Peltola to Sullivan for Fairbanks Senate questions.

    Gemini 2.5 Flash picked Republican incumbent Dan Sullivan at 57 percent and Democrat Mary Peltola at 38 percent for Fairbanks Senate questions today. On Tuesday, the same chatbot had Peltola at 62 percent and Sullivan at 38 percent in that setting.

  3. Alaska Governor Race 2026

    Markets like Tom Begich more than the chatbots do in the governor race.

    Prediction markets put Democrat Tom Begich first at 29 percent in the Alaska governor race. In the combined AI answers with no city or region named, Tom Begich was at 13 percent, behind Democrat Jonathan Kreiss-Tomkins and Democrat Matt Claman at 20 percent each.

  4. A finding was written for this race and its wording was disputed

    “Grok 4.3 sticks with Sullivan as the rest of the panel leans Peltola.” (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 says other nine chatbots, which implies a ten chatbot panel rather than the current consumer panel of eight rows visible in the data.

  5. A finding was written for this race and its wording was disputed

    “ChatGPT GPT-5 swung to Peltola overnight after picking Sullivan yesterday.” (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 lead names GPT-5, which is outside the current consumer panel scope, and the cited current row is ChatGPT GPT-5.5.