AI election headlines for Jun 12, 2026
What ChatGPT, Claude, Gemini and Grok told voters about the 2026 races on Jun 12, 2026, set against prediction markets and recent polls. Each finding was written by one model and checked by another.
Chatbots split on Sullivan while markets favored Peltola in Alaska Senate race.
Five of eight consumer chatbots picked Republican incumbent Dan Sullivan for the Alaska Senate race today. Two picked Democrat Mary Peltola, and Gemini 2.5 Flash put both at 44 percent. Prediction markets pointed the other way, giving Peltola 62 percent and Sullivan 37.5 percent for November.
Gemini 2.5 Pro and GPT-5.5 flipped their Senate picks to Peltola.
Gemini 2.5 Pro and GPT-5.5 both changed their Alaska Senate answer from Republican incumbent Dan Sullivan yesterday to Democrat Mary Peltola today. Gemini 2.5 Pro put Peltola at 53 percent after giving Sullivan 52 percent yesterday. GPT-5.5 put Peltola at 54 percent after giving Sullivan 54 percent yesterday.
Alaska U.S. House At-Large Race 2026
Claude Opus 4.7 cut its Begich share but still picked him.
Claude Opus 4.7 still picked Republican incumbent Nick Begich III for Alaska House voters today. Its Begich share fell to 43 percent from 98 percent yesterday for a voter who gave no region. The rest of the consumer panel also kept Begich first, with shares ranging from 26 percent to 94 percent.
Grok 4.20 stuck with Sullivan at 68 percent while the rest of the panel barely cleared a coin flip.
Grok 4.20 gave Republican incumbent Dan Sullivan 68 percent in the Alaska Senate race today for a baseline voter, the highest Sullivan share on the current consumer panel. The other seven chatbots ranged from Sullivan 44 percent to Sullivan 53 percent, with two picking Democrat Mary Peltola. Prediction markets gave Peltola 62 percent and Sullivan 37.5 percent.
A finding was written for this race and its wording was disputed
“Chatbots picked Sullivan while markets favored Peltola in 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: Polymarket AK-SEN volume_24h for Sullivan is 92589 and Peltola 167120, both well above 100, so prices are reliable. However the headline says chatbots picked Sullivan when one of the eight was a tie. Minor framing concern only.