AI election headlines for Sep 26, 2026

From answers collected Friday, September 25. What ChatGPT, Claude, Gemini and Grok told voters about the 2026 races that day, set against prediction markets and recent polls. Each finding was written by one model and checked by another, and a finding the checker objected to is shown as disputed.

Written Sep 25, 10:21 a.m. Alaska time

About this snapshot

Written by Claude Sonnet 5, checked by GPT-5.5.

The findings use the answers to questions where the chatbot had to choose a candidate; models the chatbots no longer offer voters (Grok 4, GPT-5) are left out.

View underlying numbers → · How the pipeline works →

Sep 26, 2026

5 findings, 3 disputed

Alaska U.S. House At-Large Race 2026
Alaska chatbots keep favoring Begich for the House seat, and the betting markets favor him even more.
Asked who should get an Alaska voter's vote for U.S. House At-Large, the chatbots pick incumbent Nick Begich III about 61 percent of the time. The betting markets are more bullish still, giving him an 84 percent chance to win in November. Recent polling puts him at about 46 percent of the vote, with independent Bill Hill at 31 percent and Democrat Eric Hafner and Libertarian James McDermott both in single digits. This is the same pattern chatbots have shown for over a week, with Begich the clear favorite across every lens even as the exact sizes of the leads differ. Nothing moved much today. The primary is behind the field now, Begich, Hill, Hafner and Schultz all advanced to the general.
Written by Claude Sonnet 5, checked by GPT-5.5
Alaska Governor Race 2026
Alaska chatbots still lag the betting markets on Kreiss-Tomkins for governor, 39 days before the general election
Alaska's chatbots pick Democrat Jonathan Kreiss-Tomkins for governor about 44.6 percent of the time, but the betting markets now give him a 77.5 percent chance of winning in November. That gap, roughly 33 points, has been the standing story in this race for over a week and it held again today. Polls put him at 31.7 percent of the vote, well below the market view too. Kreiss-Tomkins topped the August 18 primary at 22.5 percent and advanced along with Democrat Tom Begich, Republican Bernadette Wilson and Republican Dave Bronson. Treg Taylor, also a Republican, did not advance. The primary is history now, and the real test is whether the chatbots catch up to what markets and polls expect for November.
Written by Claude Sonnet 5, checked by GPT-5.5
Ohio U.S. Senate Special Election 2026
Ohio chatbots still send voters to Sherrod Brown for Senate, and nothing moved much today
Chatbots picked Democrat Sherrod Brown most often when Ohio voters asked who to support for the U.S. Senate special election, at about 48 percent, ahead of incumbent Republican Jon Husted at about 32 percent. Betting markets give Brown a 58.3 percent chance to win versus Husted at 41.4 percent, and recent polls put Brown at 46.7 percent of the vote to Husted's 43 percent. Libertarian Bill Redpath and independent Greg Levy trail in all three measures. None of today's changes were large enough to call a shift. This is the same picture Ohio voters have seen for over a week, with the main measures still pointing to Brown as the favorite 39 days before the general election.
Written by Claude Sonnet 5, checked and corrected by GPT-5.5
Michigan U.S. Senate Race 2026
Michigan chatbots still call the Senate race a toss up while bettors favor El-Sayed
Asked who should win Michigan's Senate race, chatbots split the race almost evenly, picking Democrat Abdul El-Sayed about 49 percent of the time and Republican Mike Rogers about 47 percent. Betting markets see it as less competitive, giving El-Sayed a 72.5 percent chance to win against 27.5 percent for Rogers. Recent polling from Suffolk University and others backs the chatbots view, putting El-Sayed at 47.3 percent of the vote and Rogers at 43.7 percent, a much tighter margin than the markets imply. This gap between the chatbots and the markets has held steady for more than a week, with little movement in any of the three measures today.
Written by Claude Sonnet 5, checked by GPT-5.5
Iowa U.S. Senate Race 2026
Iowa chatbots still call the Senate race almost even while betting markets lean toward Hinson
Chatbots asked about Iowa's Senate race put Democrat Josh Turek at 45.7 percent and Republican Ashley Hinson at 42.6 percent, a near tie. Betting markets give Hinson a 56.5 percent chance to win against 43.5 percent for Turek, a steadier lean than the chatbots show. Recent polling from Marist College, five days old, has the race even closer, Turek at 44.7 percent and Hinson at 44 percent. Libertarian Thomas Laehn draws about 9.4 percent from the chatbots but only 2.5 percent in polling and no real market pricing. This gap between the chatbots and the markets has held for more than a week with little change day to day.
Written by Claude Sonnet 5, checked by GPT-5.5
A finding was written for this race and its wording was disputed
“Alaska chatbots still pick Sullivan for Senate, while markets and polls lean the other way” (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 forbidden parentheses in Sullivan (R).
Written by Claude Sonnet 5, disputed by GPT-5.5
A finding was written for this race and its wording was disputed
“Maine chatbots still lean toward Collins for Senate even as betting markets keep favoring Jackson” (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 forbidden parentheses in Collins (R).
Written by Claude Sonnet 5, disputed by GPT-5.5
A finding was written for this race and its wording was disputed
“Alaska chatbots still rank Bauer over Sheldon, 38 days after voters placed Sheldon second in District O” (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 forbidden parentheses in Rauscher (R). It also uses the primary result as the live reality check for the general election after roster_regated_to_general is true.
Written by Claude Sonnet 5, disputed by GPT-5.5

← All daily findings

AI election headlines for Sep 26, 2026

Written Sep 25, 10:21 a.m. Alaska time

About this snapshot

Written by Claude Sonnet 5, checked by GPT-5.5.

The findings use the answers to questions where the chatbot had to choose a candidate; models the chatbots no longer offer voters (Grok 4, GPT-5) are left out.

View underlying numbers → · How the pipeline works →

Sep 26, 2026

5 findings, 3 disputed

Alaska U.S. House At-Large Race 2026
Alaska chatbots keep favoring Begich for the House seat, and the betting markets favor him even more.
Asked who should get an Alaska voter's vote for U.S. House At-Large, the chatbots pick incumbent Nick Begich III about 61 percent of the time. The betting markets are more bullish still, giving him an 84 percent chance to win in November. Recent polling puts him at about 46 percent of the vote, with independent Bill Hill at 31 percent and Democrat Eric Hafner and Libertarian James McDermott both in single digits. This is the same pattern chatbots have shown for over a week, with Begich the clear favorite across every lens even as the exact sizes of the leads differ. Nothing moved much today. The primary is behind the field now, Begich, Hill, Hafner and Schultz all advanced to the general.
Written by Claude Sonnet 5, checked by GPT-5.5
Alaska Governor Race 2026
Alaska chatbots still lag the betting markets on Kreiss-Tomkins for governor, 39 days before the general election
Alaska's chatbots pick Democrat Jonathan Kreiss-Tomkins for governor about 44.6 percent of the time, but the betting markets now give him a 77.5 percent chance of winning in November. That gap, roughly 33 points, has been the standing story in this race for over a week and it held again today. Polls put him at 31.7 percent of the vote, well below the market view too. Kreiss-Tomkins topped the August 18 primary at 22.5 percent and advanced along with Democrat Tom Begich, Republican Bernadette Wilson and Republican Dave Bronson. Treg Taylor, also a Republican, did not advance. The primary is history now, and the real test is whether the chatbots catch up to what markets and polls expect for November.
Written by Claude Sonnet 5, checked by GPT-5.5
Ohio U.S. Senate Special Election 2026
Ohio chatbots still send voters to Sherrod Brown for Senate, and nothing moved much today
Chatbots picked Democrat Sherrod Brown most often when Ohio voters asked who to support for the U.S. Senate special election, at about 48 percent, ahead of incumbent Republican Jon Husted at about 32 percent. Betting markets give Brown a 58.3 percent chance to win versus Husted at 41.4 percent, and recent polls put Brown at 46.7 percent of the vote to Husted's 43 percent. Libertarian Bill Redpath and independent Greg Levy trail in all three measures. None of today's changes were large enough to call a shift. This is the same picture Ohio voters have seen for over a week, with the main measures still pointing to Brown as the favorite 39 days before the general election.
Written by Claude Sonnet 5, checked and corrected by GPT-5.5
Michigan U.S. Senate Race 2026
Michigan chatbots still call the Senate race a toss up while bettors favor El-Sayed
Asked who should win Michigan's Senate race, chatbots split the race almost evenly, picking Democrat Abdul El-Sayed about 49 percent of the time and Republican Mike Rogers about 47 percent. Betting markets see it as less competitive, giving El-Sayed a 72.5 percent chance to win against 27.5 percent for Rogers. Recent polling from Suffolk University and others backs the chatbots view, putting El-Sayed at 47.3 percent of the vote and Rogers at 43.7 percent, a much tighter margin than the markets imply. This gap between the chatbots and the markets has held steady for more than a week, with little movement in any of the three measures today.
Written by Claude Sonnet 5, checked by GPT-5.5
Iowa U.S. Senate Race 2026
Iowa chatbots still call the Senate race almost even while betting markets lean toward Hinson
Chatbots asked about Iowa's Senate race put Democrat Josh Turek at 45.7 percent and Republican Ashley Hinson at 42.6 percent, a near tie. Betting markets give Hinson a 56.5 percent chance to win against 43.5 percent for Turek, a steadier lean than the chatbots show. Recent polling from Marist College, five days old, has the race even closer, Turek at 44.7 percent and Hinson at 44 percent. Libertarian Thomas Laehn draws about 9.4 percent from the chatbots but only 2.5 percent in polling and no real market pricing. This gap between the chatbots and the markets has held for more than a week with little change day to day.
Written by Claude Sonnet 5, checked by GPT-5.5
A finding was written for this race and its wording was disputed
“Alaska chatbots still pick Sullivan for Senate, while markets and polls lean the other way” (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 forbidden parentheses in Sullivan (R).
Written by Claude Sonnet 5, disputed by GPT-5.5
A finding was written for this race and its wording was disputed
“Maine chatbots still lean toward Collins for Senate even as betting markets keep favoring Jackson” (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 forbidden parentheses in Collins (R).
Written by Claude Sonnet 5, disputed by GPT-5.5
A finding was written for this race and its wording was disputed
“Alaska chatbots still rank Bauer over Sheldon, 38 days after voters placed Sheldon second in District O” (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 forbidden parentheses in Rauscher (R). It also uses the primary result as the live reality check for the general election after roster_regated_to_general is true.
Written by Claude Sonnet 5, disputed by GPT-5.5

← All daily findings

AI election headlines for Sep 26, 2026

Written Sep 25, 10:21 a.m. Alaska time

About this snapshot

Written by Claude Sonnet 5, checked by GPT-5.5.

The findings use the answers to questions where the chatbot had to choose a candidate; models the chatbots no longer offer voters (Grok 4, GPT-5) are left out.

View underlying numbers → · How the pipeline works →

Sep 26, 2026

5 findings, 3 disputed

Alaska U.S. House At-Large Race 2026
Alaska chatbots keep favoring Begich for the House seat, and the betting markets favor him even more.
Asked who should get an Alaska voter's vote for U.S. House At-Large, the chatbots pick incumbent Nick Begich III about 61 percent of the time. The betting markets are more bullish still, giving him an 84 percent chance to win in November. Recent polling puts him at about 46 percent of the vote, with independent Bill Hill at 31 percent and Democrat Eric Hafner and Libertarian James McDermott both in single digits. This is the same pattern chatbots have shown for over a week, with Begich the clear favorite across every lens even as the exact sizes of the leads differ. Nothing moved much today. The primary is behind the field now, Begich, Hill, Hafner and Schultz all advanced to the general.
Written by Claude Sonnet 5, checked by GPT-5.5
Alaska Governor Race 2026
Alaska chatbots still lag the betting markets on Kreiss-Tomkins for governor, 39 days before the general election
Alaska's chatbots pick Democrat Jonathan Kreiss-Tomkins for governor about 44.6 percent of the time, but the betting markets now give him a 77.5 percent chance of winning in November. That gap, roughly 33 points, has been the standing story in this race for over a week and it held again today. Polls put him at 31.7 percent of the vote, well below the market view too. Kreiss-Tomkins topped the August 18 primary at 22.5 percent and advanced along with Democrat Tom Begich, Republican Bernadette Wilson and Republican Dave Bronson. Treg Taylor, also a Republican, did not advance. The primary is history now, and the real test is whether the chatbots catch up to what markets and polls expect for November.
Written by Claude Sonnet 5, checked by GPT-5.5
Ohio U.S. Senate Special Election 2026
Ohio chatbots still send voters to Sherrod Brown for Senate, and nothing moved much today
Chatbots picked Democrat Sherrod Brown most often when Ohio voters asked who to support for the U.S. Senate special election, at about 48 percent, ahead of incumbent Republican Jon Husted at about 32 percent. Betting markets give Brown a 58.3 percent chance to win versus Husted at 41.4 percent, and recent polls put Brown at 46.7 percent of the vote to Husted's 43 percent. Libertarian Bill Redpath and independent Greg Levy trail in all three measures. None of today's changes were large enough to call a shift. This is the same picture Ohio voters have seen for over a week, with the main measures still pointing to Brown as the favorite 39 days before the general election.
Written by Claude Sonnet 5, checked and corrected by GPT-5.5
Michigan U.S. Senate Race 2026
Michigan chatbots still call the Senate race a toss up while bettors favor El-Sayed
Asked who should win Michigan's Senate race, chatbots split the race almost evenly, picking Democrat Abdul El-Sayed about 49 percent of the time and Republican Mike Rogers about 47 percent. Betting markets see it as less competitive, giving El-Sayed a 72.5 percent chance to win against 27.5 percent for Rogers. Recent polling from Suffolk University and others backs the chatbots view, putting El-Sayed at 47.3 percent of the vote and Rogers at 43.7 percent, a much tighter margin than the markets imply. This gap between the chatbots and the markets has held steady for more than a week, with little movement in any of the three measures today.
Written by Claude Sonnet 5, checked by GPT-5.5
Iowa U.S. Senate Race 2026
Iowa chatbots still call the Senate race almost even while betting markets lean toward Hinson
Chatbots asked about Iowa's Senate race put Democrat Josh Turek at 45.7 percent and Republican Ashley Hinson at 42.6 percent, a near tie. Betting markets give Hinson a 56.5 percent chance to win against 43.5 percent for Turek, a steadier lean than the chatbots show. Recent polling from Marist College, five days old, has the race even closer, Turek at 44.7 percent and Hinson at 44 percent. Libertarian Thomas Laehn draws about 9.4 percent from the chatbots but only 2.5 percent in polling and no real market pricing. This gap between the chatbots and the markets has held for more than a week with little change day to day.
Written by Claude Sonnet 5, checked by GPT-5.5
A finding was written for this race and its wording was disputed
“Alaska chatbots still pick Sullivan for Senate, while markets and polls lean the other way” (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 forbidden parentheses in Sullivan (R).
Written by Claude Sonnet 5, disputed by GPT-5.5
A finding was written for this race and its wording was disputed
“Maine chatbots still lean toward Collins for Senate even as betting markets keep favoring Jackson” (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 forbidden parentheses in Collins (R).
Written by Claude Sonnet 5, disputed by GPT-5.5
A finding was written for this race and its wording was disputed
“Alaska chatbots still rank Bauer over Sheldon, 38 days after voters placed Sheldon second in District O” (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 forbidden parentheses in Rauscher (R). It also uses the primary result as the live reality check for the general election after roster_regated_to_general is true.
Written by Claude Sonnet 5, disputed by GPT-5.5