Alaska U.S. House At-Large Race 2026

Five days before the primary, chatbots still point Alaska House voters to incumbent Begich.

Chatbots picked incumbent Nick Begich III, a Republican, most often in Alaska's U.S. House race, choosing him 64.7 percent of the time, while the latest poll average, now 28 days old, put Begich at 50.7 percent of the vote. There is no live prediction market for this race, so voters do not have a betting market check on the chatbot read. The chatbot panel held roughly steady from yesterday, with no meaningful new move in the race.

Alaska U.S. Senate Race 2026

Alaska Senate chatbots, markets and older polls still show a near-even Sullivan-Peltola race five days before the primary.

Chatbots picked incumbent Dan Sullivan (R) 49.0 percent of the time and Democrat Mary Peltola 48.4 percent, while markets give Peltola a 51.0 percent chance to win and Sullivan a 48.5 percent chance. The latest polling average, based on surveys whose newest poll ended 45 days ago, puts Peltola at 47.5 percent of the vote and Sullivan at 45.1 percent. None of the three views has made a clear new move since yesterday, so the useful takeaway for Alaska voters is steadiness: the major chatbots, betting markets and polls all describe a very close U.S. Senate race.

Alaska State Senate District O Race 2026

Chatbots still put incumbent George Rauscher first in State Senate District O, with no market or poll check five days before the primary

Five days before Alaska's top-4 primary, chatbots picked incumbent George Rauscher (R) most often in State Senate District O, at 45.2 percent. They picked Democrat Peter Bauer 28.9 percent of the time and Republican Ryan Sheldon 15.6 percent. There is no live prediction market and no polling in the data for this race, so voters should treat this as a read on what the chatbots are saying, not a market forecast or a voter survey.

Alaska Governor Race 2026

In the Alaska governor's race, markets and chatbots can't agree who's even in second place.

Prediction markets put Tom Begich far ahead in Alaska's governor primary at 54.5 percent, with Bernadette Wilson a distant second at 26.5 percent. But the AI chatbot panel picks Bill Walker most often, at 20.9 percent, with Tom Begich well behind at 13.9 percent. The three-way split between what markets, chatbots and the small polling sample say about the frontrunner is the more voter-relevant story than any single candidate's move.

A finding was written for this race and its wording was disputed

“Markets pulled back from Bernadette Wilson, but still rate her far above Alaska’s governor chatbots.” (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 frames Wilson as being rated 'far above' the chatbots' view, but this omits that Wilson is not even the market frontrunner (tom_begich leads markets at 54.5 percent), so presenting her market number without that context misleads on how competitive she actually is. This is a framing problem: the piece should either lead with the actual frontrunner disagreement or clearly note Wilson's rank, not frame a decline as an implicit 'far above' claim.