Trump's Next Attorney General Odds: The AI Model Verdict Vs. The Kalshi Market
Todd Blanche has run the Justice Department as acting attorney general since Pam Bondi was removed on April 2, 2026, and on August 3, 2026 the last two Republican senators blocking his confirmation dropped their objections. Kalshi turned the question of who becomes the next confirmed attorney general into a market, one yes/no contract per person, and the crowd now prices Blanche near 96 cents. We asked six AI models to score the whole field without letting any of them see a single price. They came back with the same runaway favorite, but with a little more doubt than a 96-cent price shows, and with a useful reminder that this contract does not pay out until someone is actually confirmed.
New to Kalshi event markets? Start here → how Kalshi's yes/no contracts work.
The Quick Answer
Kalshi's market makes Todd Blanche the heavy favorite at about 96 cents, treating his confirmation as nearly settled. Our price-blind, six-model panel leans the same way, blending Blanche to about 90%. The gap is small and it points the other way from most of our boards: here the models are slightly more cautious than the crowd, holding a wider tail for votes that are scheduled but not yet cast. Every other name on the board, from Lee Zeldin to Ron DeSantis to Ken Paxton, is a contingency the panel prices under 1%, and the "No one gets confirmed" outcome sits near 2%.
- Market Favorite: Todd Blanche, ~96¢
- Model Favorite: Todd Blanche, ~90% blend (same favorite, a touch less certain)
- Where The Models Diverge From The Price: they trim the favorite from ~96¢ to a ~90% blend and keep about a 10% combined tail across "No one" and the contingency field
- The Key Wrinkle Most Readers Miss: Blanche already runs the department, but the contract settles on Senate confirmation, not on who is acting
- The Panel's Logic In One Line: a sole nominee whose last two Senate obstacles just cleared is close to a lock, but "scheduled" is not "done," and a small tail is honest.
The confirmation timeline, the settlement wrinkle that keeps this market live, the full board, the dark-horse backups, and a plain-English guide to reading these prices are all below.
Want the graded result? Every number in this piece gets scored in public once Kalshi settles the market. The settled record for this board and every other lives on our Model Verdict scoreboard, and you can get the re-scored verdict when this one resolves through the email capture at the end of this article.
The Market At A Glance
| Venue | Kalshi, a CFTC-regulated event-contract exchange (18+; availability varies by state, as of August 2026) |
| The Event | "Trump's next Attorney General?" (Kalshi event KXNEXTAG-29), one yes/no contract per person |
| The Contract | Each market resolves YES only for the first NEW person confirmed by the Senate, or installed by recess appointment, as attorney general before January 20, 2029. Acting and interim service does not count |
| Window | The board's listed expected expiration is January 20, 2029, but it can resolve early the moment a confirmation happens; the live catalyst is a committee vote scheduled for August 4, 2026 |
| Prices Below As Of | August 3, 2026 (latest snapshot; prices on this page refresh during the news wave into the confirmation votes) |
These are model estimates, not predictions of fact and not financial or voting advice. Kalshi event contracts trade on a CFTC-regulated exchange; you must be 18 or older and in an eligible state to participate, and this article does not advocate for or against any nominee or any political outcome. Every number in this piece gets graded in public once the market settles, in the full graded Model Verdict scoreboard.
The Wrinkle That Keeps This Market Alive
Start with the fact that trips up most readers: Todd Blanche is already the attorney general in practice. He has run the Justice Department since April 2, 2026. So why is there still a market on who the "next" attorney general will be?
Because of how the contract settles. Per Kalshi's own rules, acting and interim officials do not count. A contract pays out only when a person is confirmed by the Senate to the office, or installed through a recess appointment. Blanche is running the department under an acting title, which does not resolve anything. His Kalshi contract flips to YES only when the Senate confirms him, or if he takes the seat by recess appointment. That is the entire reason a person who already holds the job can still trade at 96 cents instead of a settled 100.
That distinction also shapes the tail. A single failed procedural vote would not end Blanche's path, because a recess appointment remains available and the market runs all the way to January 20, 2029. As Kimi put it after reading the other models, the true failure probability "is a product of two unlikely events rather than one," which is part of why the panel stays high even while pricing in real execution risk.
What Actually Moved This Market
Three fetched facts define the board, and the third is why the price jumped this week.
- Blanche Is The Sole Active Nominee. Trump removed Pam Bondi as attorney general on April 2, 2026, with multiple outlets reporting his displeasure over the handling of files related to Jeffrey Epstein, and named Blanche acting head the same day. In June 2026 Trump formally nominated Blanche for the permanent post. No other person on the board has been nominated.
- Two Republicans Had Stalled Him, Then Cleared. In late July the Senate Judiciary Committee postponed its vote when Sens. John Cornyn and Thom Tillis withheld support over a Justice Department $1.8 billion "anti-weaponization fund." On August 3, 2026 the department agreed to end that fund, Cornyn announced a deal to advance the nomination, and Tillis said he was satisfied. Both former holdouts now back him.
- The Vote Is On The Calendar. The committee rescheduled its vote for the morning of Tuesday, August 4, 2026. Reporting indicates that if Blanche is voted out of committee, Majority Leader John Thune could set up a first floor vote around Friday, August 7 and a final confirmation vote around Saturday, August 8. Republicans hold the Senate majority, and a Cabinet confirmation needs only a simple majority.
That is the whole story of the board. The two named obstacles just cleared, the schedule is set, and the majority math favors confirmation. The models price that as close to a lock, minus a discount for the fact that no vote has actually happened yet.
Here is the short version first, then every model's raw number.
The Board
| Outcome | Status | Market | Model blend |
|---|---|---|---|
| Todd Blanche | Active nominee (currently acting AG) | ~96¢ | 90.2% |
| No One | Office stays acting through Jan 2029 | ~2¢ | 2.0% |
| Lee Zeldin | Contingency (not nominated) | ~1¢ | <0.5% |
| Ron DeSantis | Contingency (not nominated) | ~1¢ | <0.5% |
| Rudy Giuliani | Contingency (not nominated) | ~1¢ | <0.5% |
| Yaakov Roth | Contingency (not nominated) | ~1¢ | <0.5% |
| Harmeet Dhillon | Contingency (not nominated) | <1¢ | <0.5% |
| Ken Paxton | Contingency (not nominated) | <1¢ | <0.5% |
| Mike Lee | Contingency (not nominated) | <1¢ | <0.5% |
| Jeanine Pirro | Contingency (not nominated) | <1¢ | <0.5% |
| Matt Gaetz | Contingency (not nominated) | <1¢ | <0.5% |
| Eric Schmitt | Contingency (not nominated) | <1¢ | <0.5% |
| Robert Giuffra | Contingency (not nominated) | <1¢ | <0.5% |
| Ed Martin | Contingency (not nominated) | <1¢ | <0.5% |
| Ted Cruz | Contingency (not nominated) | <1¢ | <0.5% |
| Jay Clayton | Contingency (not nominated) | <1¢ | <0.5% |
| Andrew Bailey | Contingency (not nominated) | <1¢ | <0.5% |
| Ashley Moody | Contingency (not nominated) | <1¢ | <0.5% |
| Alina Habba | Contingency (not nominated) | <1¢ | <0.5% |
| Aileen Cannon | Contingency (not nominated) | <1¢ | <0.5% |
| Jeff Jensen | Contingency (not nominated) | <1¢ | <0.5% |
| Jason Miyares | Contingency (not nominated) | <1¢ | <0.5% |
| Matt Whitaker | Contingency (not nominated) | <1¢ | <0.5% |
Every Seat's Number
Below the top two outcomes, every contingency name drew only leftover "field" mass from the models, so the honest per-model view is the favorite, the "No one" outcome, and the combined remainder.
| Outcome | Market | Fable | Opus | Sonnet | GLM | Kimi | DeepSeek | Blend | Gap (blend% − mkt¢) |
|---|---|---|---|---|---|---|---|---|---|
| Todd Blanche | ~96¢ | 90.0 | 92.0 | 90.0 | 90.0 | 88.0 | 91.1 | 90.2% | −6 |
| No One | ~2¢ | 2.5 | 1.0 | 3.0 | 1.0 | 1.5 | 3.0 | 2.0% | 0 |
| Rest Of Field (21 Names) | ~7¢ | 7.5 | 7.0 | 7.0 | 9.0 | 10.5 | 5.9 | 7.8% | +1 |
These are model estimates, not predictions of fact and not financial or voting advice. The Blend column is the simple mean of the six model numbers shown in its row. The last column is a quick directional gap, the blend percentage minus the raw market price in cents; read it as direction and size, not a precise edge. Market figures are raw last-trade YES prices, and this board's raw prices sum to a little over 100 cents, so there is a small amount of spread to remove. Every one of the 21 contingency names trades at a cent or less and drew a model number under half a percent, so treat that whole tier as one shared "someone other than Blanche" bucket, not 21 distinct reads.
A note on the panel. Our standard Model Verdict lineup runs eight seats. On this run two were unavailable: the ChatGPT and Gemini seats were not installed on the machine that produced these verdicts. This board therefore runs on six models: Claude Fable, Claude Opus, Claude Sonnet, GLM, Kimi, and DeepSeek. All six scored the field on August 3, 2026 without seeing any prices, working from a fetched, dated summary of the nomination, the confirmation timeline, and the Senate math. They then read one another's anonymized reasoning and revised. The revision round barely moved the board: Opus nudged Blanche from 92 to 91 on the point that the votes are "scheduled/pending rather than done," Kimi lifted its number from 88 to 91 on the recess-appointment backstop, and the rest held near 90. Every model's number also lands in our back-tested Model Verdict ledger, where past price-blind calls are graded against settled results so the method is measured, not just asserted.
The one line that explains the whole board: the market treats Blanche's confirmation as all but done; the models treat it as a very strong favorite with a handful of votes still to be cast. Both make him the pick, but only one of them calls it 96 cents.
Todd Blanche: The Nominee Both The Market And The Models Back
Every model made Blanche the favorite, and none put him below 88%. The case is structural and current. He is the only person on the board Trump has actually nominated, both committee-level Republican holdouts resolved their objections on August 3, and the committee vote is set for the next morning. Fable summed up the read: Blanche is "the sole active nominee with both committee holdouts resolved Aug 3 and a scheduled Aug 4 committee vote, leaving only residual procedural/withdrawal" risk. DeepSeek was blunter: "Blanche's confirmation path cleared with Cornyn/Tillis backing; committee vote set Aug 4; simple majority in GOP-led Senate highly probable."
Where the panel held back was on execution. The votes are scheduled, not counted. Opus, the seat that moved most on revision, wrote that its peers "correctly emphasized that the votes are still scheduled/pending rather than done," an execution-risk point it had under-weighted by pricing its tail too thin, and it trimmed its Blanche number a point while keeping him at the top of the cluster. That is the exact difference between a 90% blend and a 96-cent price: not doubt about the favorite, but respect for the small gap between "on the calendar" and "confirmed."
"No One": What If The Seat Just Stays Acting?
The second-largest outcome on the board is not a person. It is the "No one" contract, which resolves YES only if no new attorney general is confirmed or recess-appointed before January 20, 2029, meaning the department is led by acting officials for the rest of the term. The market prices this near 2 cents and the panel blends it to about 2%, essentially agreement.
It is a small number for a reason. There is an active, Senate-majority-backed nominee with a scheduled committee vote, so the path to some confirmation is wide open. But the outcome is not zero, and it is the cleanest way to hold the tail risk that Blanche stalls and no replacement is confirmed either. The models keep it live precisely because a nomination can fail in ways that do not immediately produce a confirmed successor.
Dark Horses The Panel Won't Dismiss
Below Blanche and the "No one" line, the board is a list of contingencies: names the market would look to only if Blanche's nomination somehow collapsed. No model singled any of them out with a distinctive number, which is itself the finding here. But the order book does rank them, and that ranking is the honest content:
- Lee Zeldin, Ron DeSantis, Rudy Giuliani, And Yaakov Roth sit at the top of the contingency tier, each trading around a single cent. They are the market's rough "next in line" list, not nominees. Note that Zeldin's price fell from about 7 cents to roughly 1 cent as Blanche's rose, a clean illustration of a field collapsing into one favorite.
- Ken Paxton, Harmeet Dhillon, Jeanine Pirro, And Mike Lee form the next layer, all trading under a cent. Each is a recognizable name in Republican legal and political circles, which is why they carry a contract at all, but none has been nominated.
- The Remaining Names (from Ted Cruz to Aileen Cannon to Alina Habba) are longshot listings at the exchange's one-cent floor. The panel handed the entire 21-name contingency field about 8% combined, not to any single person.
The takeaway on the dark horses is unusual: the most honest thing to say is that there are no live dark horses right now, only a backup board that would matter if the front-runner fell.
The Traps: Rows That Look Like Signal But Are Not
Transparency matters more than a clean story, so here is where the table would mislead you:
- A Contingency Name Is Not A Candidate. Every name below Blanche has a listed contract, but none has been nominated for the job. Their prices are backup-scenario pricing, not evidence of an active bid, and the tiny model numbers next to them are leftover field mass.
- "Acting" Is Not "Confirmed." The single biggest misread on this board is assuming Blanche's contract should already be at 100 because he runs the department. It settles on confirmation, so the roughly 4-cent gap below 100 is the market's price on the confirmation actually landing, not on who is in the office today.
- A Sub-Penny Price Is Not The Same As Zero. The one-cent floor names are not measured 0% chances; they are thin-market prices on scenarios the exchange still lists. Do not read a real signal into a 0.2% blend row.
How To Read These Prices
The models can tell you where they part from the crowd, but the number in front of you still has to be understood in context. Here is how to read a market like this one, plus the honest friction of a thin, event-driven board.
Work an example. Say Blanche's contract shows a 96-cent YES price. Taken at face value, that implies about a 96% chance he is the next confirmed attorney general. To get a fair, spread-free read you normalize the whole field so every outcome's price sums to 100. On most prediction-market boards the raw prices sum to more than 100 because each contract carries its own bid/ask spread, and de-vigging pulls the favorite's number down a touch. On this board the raw prices sum to a little over 100 cents, so a fair version trims Blanche only slightly, to roughly the mid-90s. The lesson holds anyway: always check what a fair version of a price looks like before comparing it to anything, because the favorite's number is usually a little softer than the raw cents suggest. That habit is the whole idea behind converting Kalshi's cents into real odds before comparing anything.
That same discipline of comparing a number to its fair value is what line shopping and finding the +EV side is about. OddsShopper's odds screen and free expert picks cover the sports slate rather than confirmation politics, and this market itself lives on Kalshi, but the method is the same one our House control 2026 model verdict and Fed rate cuts 2026 model verdict use on very different questions, and every number here lands in our back-tested Model Verdict ledger.
A few honest points before reading too much into any number in the table above:
- Eligibility. Kalshi is a CFTC-regulated event-contract exchange, not a sportsbook. You must be 18 or older and located in an eligible state, and availability varies by state and can change.
- Liquidity And Spread. Below the front-runner this board trades thin, most contracts at a single cent or less with light volume, so those "prices" are soft and getting in or out at your number is not always possible.
- A Moving Story. This is a live confirmation fight with a committee vote scheduled and floor votes possibly days away; new events can move it, and these estimates reflect only what was known on August 3, 2026. Prices shown on this page refresh during the news wave into the votes.
- Not Advice. Everything here is analysis for information, not a recommendation to trade or a political endorsement. These are model estimates, not predictions of fact and not financial or voting advice.
FAQ
Who is favored to be Trump's next attorney general? Todd Blanche, the current acting attorney general, is the runaway favorite. President Trump nominated him for the permanent post in June 2026, and on August 3, 2026 the two Republican senators who had been blocking his committee vote, John Cornyn and Thom Tillis, dropped their objections after the Justice Department agreed to end a disputed $1.8 billion fund. Kalshi prices him near 96 cents. Our six-model AI panel agrees he leads but is slightly less certain, blending him to about 90%.
Is Todd Blanche already the attorney general? He runs the department, but only in an acting capacity. Blanche became acting attorney general on April 2, 2026 after President Trump removed Pam Bondi. Kalshi's market does not pay out on an acting role. It resolves only when a new person is confirmed by the Senate or installed by a recess appointment, which is why the contract is still live even though Blanche is already in the job day to day.
When could Todd Blanche be confirmed? The Senate Judiciary Committee rescheduled its vote for the morning of Tuesday, August 4, 2026. Reporting indicates that if he is voted out of committee, Majority Leader John Thune could move to a floor vote, with a first procedural vote possible around Friday, August 7 and a final confirmation vote possible around Saturday, August 8. Those are reported projections of the schedule, not completed votes.
How does the Kalshi next-attorney-general market settle? Kalshi lists a separate yes/no contract for each possible person. A contract resolves YES only if that person is the first NEW attorney general, confirmed by the Senate or given a recess appointment, before January 20, 2029. Acting and interim service does not count. The "No one" contract resolves YES only if no new person is confirmed or recess-appointed by that date.
Why do the AI models price Blanche below the market? The panel and the market both make Blanche the overwhelming favorite. The models simply keep a slightly larger tail, around 10% versus the market's roughly 4%, for the votes that are scheduled but not yet cast, plus the small chance of a late withdrawal or a new revelation before confirmation. Several models also noted that a failed floor vote would not end his chances, because a recess appointment remains a backstop and the market runs all the way to January 2029.
The Bottom Line
The market and the models are looking at the same nomination and reaching the same favorite with slightly different conviction. Kalshi prices Todd Blanche near 96 cents, close to settled; the panel blends him to about 90%, still the runaway pick but holding a wider tail for the votes still to come. The read to take is not that Blanche is in trouble, because a sole nominee whose last two Senate obstacles just cleared, with a scheduled vote and the majority math on his side, is exactly what a favorite looks like. It is that "scheduled" is not "confirmed," and this contract does not pay out until it is. If this method is your thing, our House control 2026 model verdict and our guide to Kalshi versus Polymarket are the natural next reads.
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $166M traded
- Rocket Classic Winner · $124M traded
- 2027 Pro Football Champion · $51M traded
- Pro Baseball Champion · $40M traded
- Bitcoin price at the end of 2026 · $30M traded
- 2027 Pro Basketball Champion · $19M traded
Every market above links to our full AI model verdict; browse them all on the OddsShopper prediction markets hub, and see how every settled call actually scored on the full graded scoreboard.
Verdicts generated August 3, 2026. Prices as of August 3, 2026. Panel: Claude Fable, Claude Opus, Claude Sonnet, GLM, Kimi, and DeepSeek, price-blind, after one revision round. These are model estimates, not predictions of fact and not financial or voting advice.
