Updated September 4, 2026
Prediction markets can change quickly. A court filing, election poll, military development or product announcement can move probabilities within hours.
That makes a snapshot of what traders are pricing right now very different from a list of events worth watching over the coming month.
In early September, some of the most interesting markets on Polymarket range from the future of sports prediction markets themselves to control of Congress, the Super Bowl and the rollout of OpenAI's newest AI model.
TL;DR: Some of the biggest Polymarket stories right now include:
Prices and probabilities can move continuously as traders buy and sell contracts.
One of the most topical Polymarket markets right now is about prediction markets themselves.
New Jersey asked the U.S. Supreme Court on September 2 to review its legal fight with Kalshi over whether states can apply their gambling laws to sports-event contracts offered on federally regulated prediction markets.
The request comes after federal appeals courts reached conflicting conclusions. The Third Circuit sided with Kalshi in its dispute with New Jersey, while the Ninth Circuit ruled on August 28 that Nevada could enforce its gaming laws against Kalshi's sports markets.
Polymarket traders currently put the probability of the Supreme Court accepting a sports-event-contract case by the end of 2026 at roughly 45%. The market has attracted close to $1 million in trading volume.
That does not mean traders are predicting how the Supreme Court would ultimately rule. The immediate question is simply whether the justices agree to hear the dispute.
For the prediction-market industry, however, even that decision could be significant. A Supreme Court case could eventually help determine where federal CFTC oversight ends and state gambling regulation begins.
The November midterms have become another major area of activity.
Rather than simply asking which party will control one chamber, Polymarket's Balance of Power market combines the House and Senate results.
As of September 4, a Democratic sweep is priced around 50%, while a Republican Senate and Democratic House sits around 36%. A Republican sweep is considerably further back.
The market has generated roughly $11 million in volume. Polymarket separately gives Democrats around a 90% probability of controlling the House, while its Senate market remains much tighter.
The pricing comes with Democrats showing stronger enthusiasm in some recent polling. A late-August Reuters/Ipsos survey found 46% of self-identified Democrats saying they were highly enthusiastic about voting in November, compared with 31% of Republicans.
That still leaves the Senate as the major source of uncertainty.
Races in states including Alaska have become increasingly important, and even relatively small changes in a handful of contests could move the balance-of-power probabilities considerably before November 3.
Sports have become a major part of prediction-market trading, and the NFL championship market already has almost $50 million in volume before the regular season has properly developed.
The Los Angeles Rams are the clear Polymarket favorite at around 18%.
Buffalo is next at roughly 9%, followed by defending champion Seattle around 8% and Baltimore close to 7%.
The separation at the top is notable.
Traditional sportsbooks are telling a similar story. BetMGM moved the Rams to approximately +500 after Aaron Donald confirmed his return, with Los Angeles also attracting a significant share of early Super Bowl futures action.
An 18% probability is still a long way from certainty, of course. More than 80% of Polymarket's probability remains distributed across the rest of the NFL.
But compared with the crowded group behind them, traders currently see Los Angeles as the team most clearly separated from the field.
The U.S.–Iran conflict remains one of the fastest-moving areas of Polymarket.
More than $25 million has now been traded on the market asking when the United States will announce an end to its naval blockade of Iran.
The probabilities have shifted noticeably since August.
Traders currently put the chance of an announcement by September 30 at only around 18%. That rises to roughly 36% by October 31 and 60% by December 31.
In other words, the market has increasingly moved toward a longer blockade.
There is a clear real-world reason for that repricing. The blockade remains in place, while Iranian oil exports have fallen sharply and diplomatic progress has remained limited. Reuters reported this week that Iranian crude loadings dropped dramatically during August as the blockade restricted the country's ability to move oil through the region.
This is exactly the type of prediction market that can change extremely quickly.
A diplomatic breakthrough could shift the timeline significantly. Another military escalation could move it in the opposite direction.
November 2028 remains more than two years away, but traders have already put extraordinary amounts of money behind the next presidential race.
Polymarket's overall 2028 presidential winner market has generated more than $700 million in trading volume.
Vice President JD Vance currently leads at around 24%, followed by Alexandria Ocasio-Cortez at roughly 13%, Jon Ossoff around 11.5% and Marco Rubio near 10%.
The important takeaway is not necessarily that traders strongly expect Vance to become president.
It is how fragmented the market remains.
Even the current favorite is priced below 25%, leaving roughly three-quarters of the total probability elsewhere.
At this stage of the cycle, those prices should be treated as a constantly changing measure of trader expectations rather than anything resembling a settled election forecast. Candidate announcements, the 2026 midterm results and the political environment over the next two years could completely reshape the market.
One of the fastest-moving technology markets is now centered on GPT-6 Astra.
OpenAI formally introduced Astra on September 3, but the initial rollout is limited rather than a full public launch.
OpenAI says access is currently rolling out to a limited number of organizations, with broader availability for ChatGPT Plus, Pro, Business and Enterprise users, as well as API customers, expected over the coming days.
That distinction has kept Polymarket's Astra release market active.
More than $1.4 million has been traded on the timing market. Traders currently put the probability of qualifying general availability by September 15 at roughly 97%, rising to around 99% by September 30.
The uncertainty here is very different from an election or sporting event.
There is no campaign, season or scheduled decision date. The market could effectively be decided by a single OpenAI product announcement.
That also makes it a good example of why prediction markets increasingly extend beyond traditional betting subjects. Traders can now put prices on product launches, technology milestones and corporate decisions almost as soon as those stories emerge.
The most interesting thing about the current Polymarket landscape is how little the biggest markets have in common.
Traders are simultaneously trying to price a Supreme Court decision, control of Congress, the NFL championship, the duration of a military blockade, an election more than two years away and the rollout of an AI model.
What connects them is uncertainty.
New information changes expectations, traders react and the market price moves with them.
That is one reason prediction markets can be useful even for people who never place a trade. Prices provide a continuously updating snapshot of how market participants are assessing uncertain future events.
For a more detailed look at the platform itself, read our full Polymarket review.
And for a more forward-looking selection rather than a snapshot of what is trending today, see our 5 Polymarket Markets to Watch in September 2026.
