AI election headlines for Jun 5, 2026
What ChatGPT, Claude, Gemini and Grok told voters about the 2026 races on Jun 5, 2026, set against prediction markets and recent polls. Each finding was written by one model and checked by another.
Gemini 2.5 Pro shifted further toward Peltola overnight in the Alaska Senate race.
Gemini 2.5 Pro picked Democrat Mary Peltola at 71 percent for a default Alaska Senate voter today, up from 62 percent yesterday. Republican incumbent Dan Sullivan fell to 29 percent. Prediction markets give Peltola a 61 percent chance to win in November. Gemini 2.5 Pro now sits above the market on Peltola by 10 percentage points.
Gemini 2.5 Flash flipped to Sullivan in the Senate race, right as markets still favored Peltola.
Gemini 2.5 Flash picked Republican incumbent Dan Sullivan at 52 percent for a default Alaska Senate voter today, after putting Democrat Mary Peltola at 62 percent yesterday. Prediction markets put Peltola at 61 percent and Sullivan at 39 percent.
Gemini 2.5 Flash gave Kreiss-Tomkins 62 percent for governor, while markets left him at 16 percent.
Gemini 2.5 Flash picked Democrat Jonathan Kreiss-Tomkins at 62 percent for a default governor voter today. Prediction markets put him at 16 percent and market leader Democrat Tom Begich at 29 percent.
A finding was written for this race and its wording was disputed
“Both Grok models picked Republican Nick Begich for every Alaska House voter 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 title overstates the tested scope. The data cover tested voter profiles and regions, not every Alaska House voter.
A finding was written for this race and its wording was disputed
“The AI panel keeps picking Kreiss-Tomkins for governor as markets back Tom Begich.” (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 names ChatGPT GPT-5 even though the current consumer panel excludes the legacy GPT-5 slot.