AI election headlines for Sep 2, 2026

What ChatGPT, Claude, Gemini and Grok told voters about the 2026 races on Sep 2, 2026, 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.

Generated 2026-09-02 18:16:21 UTC

About this snapshot

Written by openai:gpt-5.5, checked by anthropic:claude-sonnet-5.

Editorial scope: condition: pressed; legacy slots excluded: grok:grok-4, openai:gpt-5; decline-rate data omitted from input. Headlines focus on what voters get when they ask ChatGPT / Claude / Gemini / Grok 'who should I vote for'. Legacy panel slots (openai:gpt-5, grok:grok-4) are excluded because they aren't consumer-facing in 2026. Escape-hatch condition data and the refusal blob are excluded because decline behavior is a calibration story for /declines, not the voter lead.

View underlying numbers → · How the pipeline works →

Sep 2, 2026

3 findings, 1 disputed

Alaska Governor Race 2026
Chatbots picked Kreiss-Tomkins in Alaska's governor primary, but missed the order behind him.
Alaska voters decided the governor primary on August 18. Chatbots picked Democrat Jonathan Kreiss-Tomkins more than anyone else, at 15.8 percent, and voters also put him first, at 22.5 percent. The match stops there. This was a top four primary, so order decided who reached November. The chatbots put Republican Treg Taylor ahead of Republican Bernadette Wilson and Republican Dave Bronson, but voters put Wilson third at 10.0 percent and Bronson fourth at 8.1 percent, advancing both. Taylor finished fifth at 7.3 percent. The count remains unofficial, with 401 of 403 precincts reporting. The biggest miss was Democrat Thomas Scott Begich, whom chatbots picked at 9.5 percent while voters put him second at 20.8 percent.
Written by GPT-5.5, checked and corrected by Claude Sonnet 5
Alaska U.S. House At-Large Race 2026
Chatbots matched the order in Alaska's House primary, but overstated Begich's strength
Alaska voters decided the U.S. House primary on August 18. Chatbots picked incumbent Nick Begich III, Republican, first at 65.8 percent. Voters also put Begich first, at 44.6 percent, and he advanced. The count remains unofficial, with 401 of 403 precincts reporting. The chatbots had the field in the same order as voters among the candidates they were asked about, but they overstated Begich's support by 21.2 points. In this top four primary, the ranking decides who reaches the November ballot.
Written by GPT-5.5, checked and corrected by Claude Sonnet 5
Alaska U.S. Senate Race 2026
Chatbots still pick Sullivan, but Alaska voters put Peltola first in the Senate primary
Alaska voters decided the Senate primary on August 18. Chatbots picked Republican incumbent Dan Sullivan more often than Democrat Mary Peltola, at 51.7 percent to 45.9 percent. Voters put Peltola first at 49.5 percent and Sullivan second at 41.4 percent, and both advanced in the top four primary. The count remains unofficial, with 401 of 403 precincts reporting. The chatbots had the two candidates in reversed order.
Written by GPT-5.5, checked and corrected by Claude Sonnet 5
A finding was written for this race and its wording was disputed
“Chatbots moved toward Rauscher in District O but still rank Bauer above Sheldon after voters did not” (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 framing mixes present-tense day-over-day AI movement language ('moved toward', 'narrowed', 'the new movement is') with a result that voters have already decided. This reads as if the chatbot shift is racing toward or responding to the vote outcome, which is misleading since the election was 15 days prior and the AI panel move happened only between yesterday and today, unrelated to the vote count. The rules require past-tense framing for decided races and caution against conflating AI day-over-day movement with the vote-based reality check. The lead should either be purely about the AI day-over-day movement (a live, ongoing metric) as its own separate finding, or purely about the miss versus the settled result in past tense, not blended together in a way that implies the AI panel is 'catching up' to the election outcome.
Written by GPT-5.5, disputed by Claude Sonnet 5

← All daily findings

AI election headlines for Sep 2, 2026

Generated 2026-09-02 18:16:21 UTC

About this snapshot

Written by openai:gpt-5.5, checked by anthropic:claude-sonnet-5.

View underlying numbers → · How the pipeline works →

Sep 2, 2026

3 findings, 1 disputed

Alaska Governor Race 2026
Chatbots picked Kreiss-Tomkins in Alaska's governor primary, but missed the order behind him.
Alaska voters decided the governor primary on August 18. Chatbots picked Democrat Jonathan Kreiss-Tomkins more than anyone else, at 15.8 percent, and voters also put him first, at 22.5 percent. The match stops there. This was a top four primary, so order decided who reached November. The chatbots put Republican Treg Taylor ahead of Republican Bernadette Wilson and Republican Dave Bronson, but voters put Wilson third at 10.0 percent and Bronson fourth at 8.1 percent, advancing both. Taylor finished fifth at 7.3 percent. The count remains unofficial, with 401 of 403 precincts reporting. The biggest miss was Democrat Thomas Scott Begich, whom chatbots picked at 9.5 percent while voters put him second at 20.8 percent.
Written by GPT-5.5, checked and corrected by Claude Sonnet 5
Alaska U.S. House At-Large Race 2026
Chatbots matched the order in Alaska's House primary, but overstated Begich's strength
Alaska voters decided the U.S. House primary on August 18. Chatbots picked incumbent Nick Begich III, Republican, first at 65.8 percent. Voters also put Begich first, at 44.6 percent, and he advanced. The count remains unofficial, with 401 of 403 precincts reporting. The chatbots had the field in the same order as voters among the candidates they were asked about, but they overstated Begich's support by 21.2 points. In this top four primary, the ranking decides who reaches the November ballot.
Written by GPT-5.5, checked and corrected by Claude Sonnet 5
Alaska U.S. Senate Race 2026
Chatbots still pick Sullivan, but Alaska voters put Peltola first in the Senate primary
Alaska voters decided the Senate primary on August 18. Chatbots picked Republican incumbent Dan Sullivan more often than Democrat Mary Peltola, at 51.7 percent to 45.9 percent. Voters put Peltola first at 49.5 percent and Sullivan second at 41.4 percent, and both advanced in the top four primary. The count remains unofficial, with 401 of 403 precincts reporting. The chatbots had the two candidates in reversed order.
Written by GPT-5.5, checked and corrected by Claude Sonnet 5
A finding was written for this race and its wording was disputed
“Chatbots moved toward Rauscher in District O but still rank Bauer above Sheldon after voters did not” (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 framing mixes present-tense day-over-day AI movement language ('moved toward', 'narrowed', 'the new movement is') with a result that voters have already decided. This reads as if the chatbot shift is racing toward or responding to the vote outcome, which is misleading since the election was 15 days prior and the AI panel move happened only between yesterday and today, unrelated to the vote count. The rules require past-tense framing for decided races and caution against conflating AI day-over-day movement with the vote-based reality check. The lead should either be purely about the AI day-over-day movement (a live, ongoing metric) as its own separate finding, or purely about the miss versus the settled result in past tense, not blended together in a way that implies the AI panel is 'catching up' to the election outcome.
Written by GPT-5.5, disputed by Claude Sonnet 5

← All daily findings

AI election headlines for Sep 2, 2026

Generated 2026-09-02 18:16:21 UTC

About this snapshot

Written by openai:gpt-5.5, checked by anthropic:claude-sonnet-5.

View underlying numbers → · How the pipeline works →

Sep 2, 2026

3 findings, 1 disputed

Alaska Governor Race 2026
Chatbots picked Kreiss-Tomkins in Alaska's governor primary, but missed the order behind him.
Alaska voters decided the governor primary on August 18. Chatbots picked Democrat Jonathan Kreiss-Tomkins more than anyone else, at 15.8 percent, and voters also put him first, at 22.5 percent. The match stops there. This was a top four primary, so order decided who reached November. The chatbots put Republican Treg Taylor ahead of Republican Bernadette Wilson and Republican Dave Bronson, but voters put Wilson third at 10.0 percent and Bronson fourth at 8.1 percent, advancing both. Taylor finished fifth at 7.3 percent. The count remains unofficial, with 401 of 403 precincts reporting. The biggest miss was Democrat Thomas Scott Begich, whom chatbots picked at 9.5 percent while voters put him second at 20.8 percent.
Written by GPT-5.5, checked and corrected by Claude Sonnet 5
Alaska U.S. House At-Large Race 2026
Chatbots matched the order in Alaska's House primary, but overstated Begich's strength
Alaska voters decided the U.S. House primary on August 18. Chatbots picked incumbent Nick Begich III, Republican, first at 65.8 percent. Voters also put Begich first, at 44.6 percent, and he advanced. The count remains unofficial, with 401 of 403 precincts reporting. The chatbots had the field in the same order as voters among the candidates they were asked about, but they overstated Begich's support by 21.2 points. In this top four primary, the ranking decides who reaches the November ballot.
Written by GPT-5.5, checked and corrected by Claude Sonnet 5
Alaska U.S. Senate Race 2026
Chatbots still pick Sullivan, but Alaska voters put Peltola first in the Senate primary
Alaska voters decided the Senate primary on August 18. Chatbots picked Republican incumbent Dan Sullivan more often than Democrat Mary Peltola, at 51.7 percent to 45.9 percent. Voters put Peltola first at 49.5 percent and Sullivan second at 41.4 percent, and both advanced in the top four primary. The count remains unofficial, with 401 of 403 precincts reporting. The chatbots had the two candidates in reversed order.
Written by GPT-5.5, checked and corrected by Claude Sonnet 5
A finding was written for this race and its wording was disputed
“Chatbots moved toward Rauscher in District O but still rank Bauer above Sheldon after voters did not” (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 framing mixes present-tense day-over-day AI movement language ('moved toward', 'narrowed', 'the new movement is') with a result that voters have already decided. This reads as if the chatbot shift is racing toward or responding to the vote outcome, which is misleading since the election was 15 days prior and the AI panel move happened only between yesterday and today, unrelated to the vote count. The rules require past-tense framing for decided races and caution against conflating AI day-over-day movement with the vote-based reality check. The lead should either be purely about the AI day-over-day movement (a live, ongoing metric) as its own separate finding, or purely about the miss versus the settled result in past tense, not blended together in a way that implies the AI panel is 'catching up' to the election outcome.
Written by GPT-5.5, disputed by Claude Sonnet 5