• Missouri's Republican map returned to the US Supreme Court again, but with ballots already printed and voting underway on the old lines, VICO sees almost no chance the new map is used.

• The Court has already issued multiple orders in the case, and VICO argues the justices are unlikely to reverse course this close to the election, especially with votes already being cast on the old map.

• With the old map in place, the seat it was drawn to flip stays blue: Cleaver's Kansas City district is safe.

• The new map would have protected Ann Wagner while cracking Cleaver's seat. Without it, Wagner is exposed in a competitive district.

• The gerrymander rebounds on its own party: the seat meant to flip stays Democratic, while the Republican it was meant to protect is now the Missouri race most likely to change hands.

• • Little replaced Angie Craig, who held this swing seat from 2018, and VICO worries about his progressive platform in a traditionally moderate district.

• • The model sees real danger from her progressive flank: the primary challenger who refused to endorse her took 16% in the open primary, letting her Republican rival finish first. VICO worries those voters may stay home, opening the door to a Republican win.

• • In a Democrat-vs-Democrat general, VICO backs Matsui to near certainty on incumbency and money, while Kalshi is more bullish on progressive challenger Mai Vang.

• • The district is competitive, but VICO thinks Bouchard's dominant fundraising edge matters more than the markets do.

• • Markets favor progressive challenger William Lawrence, but VICO sees his record, including support for defunding the police, as a liability against the incumbent.

• The model still has Democrats as clear favorites to win the House, just by a smaller margin than the prediction markets.

• The largest divergences all point the same way: VICO is less bullish on progressive candidates than the markets, and more concerned about a Republican fundraising edge in a few key races.

• On Tuesday night, the final primary elections were held, officially cementing the general election slate for all 50 states.

• This section is an overview of the VICO model's projections for both the House and the Senate.

• Democrats are comfortably favored to retake the chamber, though VICO is more restrained than the betting markets, and it gives better-than-even odds that the majority will be at least 10 seats.

• A Democratic majority in the House would effectively block any non-bipartisan Trump legislation. Importantly, a victory with a strong majority would strengthen Democratic leadership and weaken the influence of the party's flanks who may break with leadership on key votes.

• The Senate is a genuine toss-up, and the tension is simple: Republicans are defending an unpopular president, but Democrats have a rough map and need to flip four seats for a majority.

• 43% Ken Paxton (R) wins Texas. Versus Polymarket at 44%.

• 34%: Dan Sullivan (R) wins Alaska. Versus Polymarket at 35%.

• 54%: Abdul El-Sayed (D) wins Michigan. Versus Polymarket at 62%.

• 41%: Troy Jackson (D) wins Maine. Versus Polymarket at 69%.

• 47%: Josh Turek (D) wins Iowa. Versus Polymarket at 41%.

• 63%: Pete Ricketts (R) wins Nebraska. Versus Polymarket at 68%.

• 64%: Sherrod Brown (D) wins Ohio. Versus Polymarket at 56%.

• 72%: Roger Marshall (R) wins Kansas. Versus Polymarket at 75%.

• • The VICO model has moved sharply towards Democrats, crossing above 50% for the first time, up from 30% in late July.

• The VICO model expects a Democratic victory in the House, and argues the Senate is a toss-up with a very slight Democratic lean. A Senate victory would allow Democrats to block any Supreme Court picks and bog down the administration with investigations.

• Since early August, VICO stayed consistently low on the CLARITY Act, the bill meant to split crypto oversight between the SEC and CFTC, never putting passage above 7% even as the industry treated it as alive.

• The CLARITY Act was Congress’s attempt to write the first comprehensive federal rules for crypto markets; most importantly, defining which digital assets fall under the SEC and which fall under the CFTC. 

• Yesterday the act failed a Senate vote, all but confirming it is dead until the next Congress, where a version would have to be introduced from scratch.

• Multiple crypto heavy stocks fell hard after the bill’s failure.

• The legislative path stays narrow at 30%, because VICO reads the bill's problem as ideological, not procedural. The fight over how much ground to cede to crypto does not reset with a new Congress, so it likely carries over intact.

• That pushes the action to the agencies, where VICO gives it 62% that the SEC and CFTC issue joint guidance or open a rulemaking before the end of 2027, doing by rule what Congress would not do by law. But rules can only go so far: the agencies can clarify what counts as a security, not hand the CFTC the spot-market authority only a statute can grant. The likely future is an administrative patch that falls short of what Congressional legislation could achieve.

• The edge was timing: VICO held under 7% for six weeks while the bill was still written up as alive. 

• The market was the one caught off guard, sending Coinbase and Circle both down about 10% on the vote.

• Last week, Republicans held a midterm convention in Texas, which served in part to fundraise for Senate candidate Ken Paxton who is on the ropes in early polling.

• The VICO model treats Paxton as a slight underdog in both ad spending, and election odds.

• Democrat James Talarico has raised a record-setting $70 million in direct donations, compared to Paxton's total around $15 million.

• In an attempt to keep Paxton alive, Republicans pledged tens of millions in PAC spending.

• But at 48%, the VICO model argues Talarico's continued grassroots support is likely to keep pace with Republican PAC spending, possibly outspending pro-Paxton forces.

• Talarico's donor network is likely to remain strong. The VICO model argues the Democrat's strong polling numbers will continue to energize donors.

• Without a fundraising advantage, Paxton will have to turn his polling deficit around by other means, leaving Talarico a slight favorite in a close race.

• As states like California pass large AI legislation packages, the federal government has remained mostly silent. 

• The VICO model expects the inaction to extend through 2027.

• The single likeliest to pass is not an AI-governance law at all, but a national-security one aimed at China, at 42%.

• Deepfake and likeness protections have a middling shot, helped by bipartisan support and concrete-harm framing, at 30%. 

• The ambitious bills are effectively dead on arrival, at 9% and 6% respectively: a preemption bill that would create federal uniformity over state AI laws, and a measure criminalizing election deepfakes.

• In the federal government's absence, expect a patchwork of different laws to be passed by different states, each taking the problem into their own hands.

  • 70%: The probability that at least five states pass major AI regulatory frameworks by the end of 2027.

• For AI firms, that means complying with a large web of legal codes, and ensuring nationwide products comply with the strictest statewide guidelines.

• House Democrats voted down a Republican amendment to lock the Supreme Court at nine justices, technically keeping the door to expansion open.

• The VICO model reads the whole fight as theater, since neither expansion nor a formal lock is anywhere close to happening.

• According to the model, expansion is a long shot at 13% even if Democrats controlled both houses and the presidency, because moderate members view the action as setting a dangerous precedent for more expansion in the future.

• A constitutional lock is even harder, effectively a non-starter at just 1%. Amending the Constitution takes a two-thirds supermajority in both chambers and ratification by 38 states, an impossible bar in a country this divided, hence a near-zero reading.

• The size of the Court will stay a live political talking point, useful for fundraising and base energy on both sides.

• But the model sees almost no chance it meaningfully changes, so the votes are a message, not a mechanism. We are calling this story noise.

California's legislature sent Governor Newsom a wave of AI bills, and he has until September 30 to act. With no pocket veto in California, an affirmative veto is the only way any of these dies. Newsom has blocked tech regulation he saw as bad for the state before, but broad political support and his own presidential ambitions cut the other way, and the VICO model expects these four to become law.

The model sees every bill more likely than not to survive. Even where big tech is lobbying hard, public appetite for regulating AI is strong, and a governor with national ambitions has little reason to spike popular child-safety and worker-protection bills. The support shows in the margins: SB 867's toy chatbot ban cleared the legislature 39-0. That is why veto odds sit low across the board. Even SB 1119, a chatbot child-safety bill which OpenAI CEO Sam Altman reportedly contacted Governor Newsom directly about, sits at just 32%. But passage is not the end: if all four become law, the model puts the odds of a DOJ challenge within six months at 52%.

The country's largest state economy, and home to most tech companies, is about to become the first major regulator of AI, and the costs are concrete. Social platforms would build separate under-16 products stripped of the feeds and autoplay that drive ad revenue, toymakers would pull chatbot products, employers would rip out emotion-tracking tools, and chatbot makers would carry new safety and liability exposure. For consumer-facing AI and platform names, that is real margin pressure to price in now, especially since California often sets the national standard. But litigation is likely to test whether or not the rules stand.