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

AI election headlines for Sep 29, 2026

What ChatGPT, Claude, Gemini and Grok told voters about the 2026 races on Sep 29, 2026, set against prediction markets and recent polls. Each finding was written by one model and checked by another.

  1. Alaska Governor Race 2026

    Alaska chatbots still pick Kreiss-Tomkins for governor, but they lag well behind what the betting markets expect.

    Asked who should win Alaska's governor race, chatbots picked Democrat Jonathan Kreiss-Tomkins about 52 percent of the time, but betting markets now give him a 77.5 percent chance to win in November. Polls put him at 31.7 percent of the vote, in a survey that finished September 17. None of today's changes in the AI, market, or poll numbers were large enough to count as a real shift. Kreiss-Tomkins led August 18's top four primary with 22.5 percent, a result that no longer measures the general election field.

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

    Alaska chatbots still favor incumbent Begich for the House seat, and the betting markets are even more sold on him.

    Asked who should win Alaska's U.S. House At-Large seat, chatbots pick incumbent Begich, a Republican, about 62 percent of the time. Betting markets go further, giving him an 84.5 percent chance to win in November, while a poll last updated 18 days ago put him at 45.7 percent of the vote. Bill Hill, an independent, trails in the chatbot answers at 16.1 percent even though that same poll had him at 31 percent. Begich led the August primary with 44.6 percent and Hill was second with 32.5 percent. Nothing moved much today.

  3. Ohio U.S. Senate Special Election 2026

    Ohio chatbots still send voters to Sherrod Brown for Senate, and the race is holding steady.

    Ohio voters asking chatbots about the special Senate race are still pointed to Democrat Sherrod Brown, whom the AI panel picks about 45 percent of the time, ahead of incumbent Jon Husted, a Republican, at about 35 percent. Betting markets agree, giving Brown roughly a 59 percent chance to win against Husted's 42 percent. The recent polling average, last updated 13 days ago, has Brown at 46.7 percent to Husted's 43 percent. Libertarian Bill Redpath and independent Greg Levy remain far back in every measure. Nothing moved much today, with 35 days left before the November 3 general election.

  4. Alaska U.S. Senate Race 2026

    Alaska chatbots still lean to Sullivan for Senate, while betting markets keep favoring Peltola.

    Chatbots asked who should get an Alaska voter's pick for Senate favor incumbent Dan Sullivan about 54 percent of the time, ahead of Democrat Mary Peltola at 40 percent. Betting markets see it the other way, giving Peltola a 72.5 percent chance to win in November against Sullivan's 26.5 percent. Recent polling is closer, with Peltola at 43.7 percent and Sullivan at 42.7 percent. This gap between the chatbots and the markets has held for more than a week with little change today. The primary on August 18 put Peltola first with 49.5 percent to Sullivan's 41.4 percent, but that vote decided who advanced to November, not who wins it, so it is not the yardstick now.

  5. Michigan U.S. Senate Race 2026

    Michigan chatbots still call the Senate race closer than betting markets do

    Voters in Michigan asking chatbots about the U.S. Senate race get an almost even split. The AI panel puts Democrat Abdul El-Sayed at 50 percent and Republican Mike Rogers at 46.5 percent. Betting markets see it as less competitive, giving El-Sayed a 72.5 percent chance to win against 27 percent for Rogers, and markets moved toward El-Sayed again today. Recent polling from Suffolk University and others backs the chatbots' closer read, with El-Sayed at 47.3 percent and Rogers at 43.7 percent among voters surveyed. This gap between what the chatbots say and what bettors expect has held for more than a week with no sign of narrowing.

  6. Texas U.S. Senate Race 2026

    Texas chatbots still lean toward Talarico for Senate, though less than the betting markets do

    Chatbots asked by Texas voters about the U.S. Senate race picked Democrat James Talarico about 46 percent of the time, ahead of Republican Ken Paxton at 38.8 percent and Libertarian Ted Brown far back. Betting markets are more bullish on Talarico, giving him a 59.5 percent chance to win versus 40.5 percent for Paxton. Recent polling, last updated September 19 by Texas Southern University and others, shows the two nearly tied, Talarico at 46.7 percent and Paxton at 45.7 percent. All three measures have favored Talarico for more than a week and nothing moved much today, so this remains a steady story rather than new news.

  7. Iowa U.S. Senate Race 2026

    Iowa chatbots keep splitting the Senate race almost evenly while betting markets still lean toward Hinson.

    Ask a chatbot who should get your vote in Iowa's Senate race and the panel comes back nearly split, picking Democrat Josh Turek about 47.7 percent of the time and Republican Ashley Hinson about 42.6 percent. Betting markets see it differently, giving Hinson a 59.5 percent chance to win against Turek's 40.5 percent. Polls, most recently from Marist College nine days ago, show the two effectively tied, 44.7 to 44 percent. Libertarian Thomas Laehn draws little support in any measure. Nothing moved much today, and this gap between the chatbots and the markets has held steady for more than a week.

  8. Alaska State Senate District O Race 2026

    Six weeks after Alaska's District O primary, chatbots still rank Bauer over Sheldon for the general election

    Asked about State Senate District O, chatbots pick incumbent George Rauscher, a Republican, most often, at 46.4 percent. That matches August 18, when he led the primary with 38.8 percent of the vote. But the chatbots still put Democrat Peter Bauer ahead of Republican Ryan Sheldon, at 28.9 percent to 16.2 percent. Voters had it the other way in the primary order, putting Sheldon second with 30.6 percent of the vote to Bauer's 30.6 percent. All three advanced to the November 3 general election. There is no live betting market or poll for this race to check the numbers against.

  9. Maine U.S. Senate Race 2026

    Maine chatbots still favor Collins for Senate even as betting markets lean toward Jackson.

    Voters in Maine asking chatbots about the Senate race are still steered toward incumbent Susan Collins about 52 percent of the time, while the betting markets give Democrat Troy Jackson a stronger chance, near 60 percent. The polling average, last updated 14 days ago, puts Jackson slightly ahead, 47 percent to 45.7 percent. That gap between what the chatbots say and what the markets show has held steady for more than a week, with little movement in any of the three measures today. The standing disagreement remains the notable fact of this race as the general election on November 3 approaches, 35 days out.

  10. Alaska Governor Race 2026

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

    “Alaska chatbots still pick Kreiss-Tomkins for governor, but they lag well behind what the betting markets expect.” (the wording our summarizing model proposed)

    Our summarizing model wrote a finding about this race today and our checking model objected to how it was worded, so it was not published as written. We show the disagreement rather than hiding it: the two models checking each other is the audit. The objection: The body is over the 120 word target. The numbers are correct, so this needs trimming rather than rejection.

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

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

    “Alaska chatbots still favor incumbent Begich for the House seat, and the betting markets are even more sold on him.” (the wording our summarizing model proposed)

    Our summarizing model wrote a finding about this race today and our checking model objected to how it was worded, so it was not published as written. We show the disagreement rather than hiding it: the two models checking each other is the audit. The objection: The body uses forbidden parentheses around R and is over the 120 word target. The fixed body removes the parentheses and trims context.

  12. Ohio U.S. Senate Special Election 2026

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

    “Ohio chatbots still send voters to Sherrod Brown for Senate, and the race is holding steady” (the wording our summarizing model proposed)

    Our summarizing model wrote a finding about this race today and our checking model objected to how it was worded, so it was not published as written. We show the disagreement rather than hiding it: the two models checking each other is the audit. The objection: The body uses forbidden parentheses around R. It also describes the 46.7 percent and 43 percent figures as polling from Trafalgar Group, but those are the three poll average, with Trafalgar only the latest pollster.

  13. Alaska State Senate District O Race 2026

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

    “Six weeks after Alaska's District O primary, chatbots still rank Bauer over Sheldon for the general election” (the wording our summarizing model proposed)

    Our summarizing model wrote a finding about this race today and our checking model objected to how it was worded, so it was not published as written. We show the disagreement rather than hiding it: the two models checking each other is the audit. The objection: The body uses forbidden parentheses around R. The claim that the chatbots have not closed the gap in over a month is not supported by the supplied movement block or seven day summary, so it was removed.

  14. Maine U.S. Senate Race 2026

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

    “Maine chatbots still favor Collins for Senate even as betting markets and polls lean toward Jackson” (the wording our summarizing model proposed)

    Our summarizing model wrote a finding about this race today and our checking model objected to how it was worded, so it was not published as written. We show the disagreement rather than hiding it: the two models checking each other is the audit. The objection: The title leans on a polling gap of only 1.3 points, and the latest poll is 14 days old. The fixed version treats the poll as older context rather than today's lead.