AI election headlines for Jun 29, 2026
What ChatGPT, Claude, Gemini and Grok told voters about the 2026 races on Jun 29, 2026, set against prediction markets and recent polls. Each finding was written by one model and checked by another.
Alaska U.S. House At-Large Race 2026
Grok 4.3 backs Nick Begich across the Alaska House regions tested.
Grok 4.3 picked incumbent Republican Nick Begich III for the Alaska House race in every region tested today, including Western Alaska, the Matanuska Susitna Valley, and Fairbanks. It gave Begich 100 percent of its picks for a default voter and 71 percent for a Western Alaska voter. Claude Sonnet 4.6 split evenly between Begich and Democrat John Brendan Williams for a default voter.
Claude Sonnet 4.6 swung hard toward Sullivan for Fairbanks Senate voters.
Claude Sonnet 4.6 gave Republican incumbent Dan Sullivan 62 percent of its Alaska Senate picks for a Fairbanks voter today, up from 20 percent yesterday. Mary Peltola fell to 38 percent today from 60 percent yesterday in the same chatbot and region.
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 says chatbots recommend candidates, which breaks the framing rule. The body also uses hyphenated wording in AI-versus-market when the published text is supposed to use periods and commas only.
A finding was written for this race and its wording was disputed
“Gemini 2.5 Flash splits its Alaska Senate pick by region, unlike every other chatbot.” (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 to every other chatbot is false. Claude Sonnet 4.6 also splits its top Senate pick by region. The body also uses punctuation beyond periods and commas.
A finding was written for this race and its wording was disputed
“Grok 4 picks Republican Nick Begich for the Alaska House race across every region tested.” (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 headline names Grok 4 as the lead finding. The current consumer panel slot is Grok 4.3.
A finding was written for this race and its wording was disputed
“Claude Sonnet 4.6 flipped its Alaska Senate pick from yesterday to 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 says Claude Sonnet 4.6 is one of the most widely used chatbots, which is not supported by the supplied data.