A few weeks ago, Polymarket approached a nonprofit newsroom called More Perfect Union. They wanted to collaborate – specifically, to have their odds integrated into the newsroom’s reporting. The idea was that prediction market probabilities would give readers better insight into the issues being covered.
More Perfect Union did what any honest journalist does before entering a partnership. They looked at the business model first.
What they found, they published instead.
The video has had 3 million views at the time I am writing this. It is worth 21 minutes of your time. But if you work in the licensed gambling industry – and particularly if you have spent any of the last decade building compliance architecture under regulatory pressure – the specific things it exposes will produce a feeling you will recognise. Not surprise. Something closer to the feeling you get when someone else gets away with exactly what you were told you couldn’t do.
What the video actually shows
More Perfect Union’s starting point is the claim both platforms lead with. Polymarket’s CEO Shayne Coplan: “There are billions of people who could find value in this. A diverse group of market participants is more accurate than any given expert. It’s the most accurate thing we have as mankind right now.” Kalshi’s pitch is similar – prediction markets “enhance democracy,” “increase voter engagement,” represent “the best mechanism to get more truth.”
You might also have seen their ads. The ones promising you’ll make rent. The ones about generational wealth. Kalshi’s CEO Tarek Mansour, promoting a $1 billion March Madness bracket offer, told his audience: “90 minutes left to attempt generational wealth.” This is the CEO of a financial platform, speaking to working people who are withdrawing from retirement accounts at record rates, telling them their grandchildren’s future is 90 minutes away.
The More Perfect Union team then looked at who was actually making money.
On Polymarket, 0.04% of traders captured 70% of the profits. The Wall Street Journal’s parallel investigation, published the same week, found 0.1% of accounts took 67% of all profits. Two thousand accounts made nearly half a billion dollars since 2022. The typical user sits on a loss between one dollar and one hundred dollars.
The who behind those winning accounts is not a mystery. Susquehanna International Group – one of the largest options market-making firms in the world – trades hundreds of millions per week through Kalshi. Jump Trading took equity stakes in both platforms. These are not curious retail participants running a better thesis about the election. They are institutions with algorithmic speed advantages the retail user cannot close by reading more carefully.
The insider trading problem no one is solving
The video’s second major finding is the one that should concern regulators most, and the one that gets the least attention in the breathless coverage of prediction markets as a financial innovation.
Senator Chris Murphy put it plainly when More Perfect Union sat down with him: “When you go on Polymarket or Kalshi you assume there’s some system in place that is assuring that the bet you are making is not rigged. But all of the bets on government action are rigged because somebody in government knows the outcome.”
Think about the structure of that problem. A prediction market on whether a specific piece of legislation passes, whether a regulatory decision goes one way or another, whether a government contract gets awarded – in each of these cases, there are people who know the answer before the market resolves. Not one person. Tens of thousands of people. Congressional staff, lobbyists, agency officials, their families, their contacts. An expert on gambling industry regulation interviewed in the video described it directly: someone who works at Spotify just has to tell a friend who’s going to have the most streams today. “How are you going to catch that?”
In the licensed gambling industry, the response to a finding like this would be immediate and extensive. The regulator would want to know what controls were in place, what surveillance was running, what the platform’s responsibility was for the integrity of its markets. The answer “we know who you are, but we can’t definitively prove insider trading” would not be considered an adequate compliance framework.
For Polymarket and Kalshi, it is currently the only answer on offer.
The reason they call it something other than gambling
This is the part of the video that the licensed gambling industry should be paying close attention to.
Prediction markets call their products “event contracts.” They describe themselves as financial platforms. They appear before Congress arguing that they are investment vehicles. The regulatory distinction matters enormously, because gambling regulation – player protection requirements, licensing costs, AML obligations, safer gambling frameworks – does not apply to financial products in the same way.
The video points to an early Kalshi ad that claimed: “Sports betting is legal in all 50 states.” It is not. What Kalshi offers is legal in all 50 states, because it is not classified as sports betting. It is classified as something else. The ad was the slip – the moment the marketing said what the legal architecture was designed to obscure.
The business logic is transparent once you see it. If prediction markets are gambling, they face the full weight of gambling regulation: licensing requirements, player protection obligations, advertising standards, AML frameworks. If they are financial instruments, they face a different set of requirements – still significant, but not the specific requirements the gambling industry spent fifteen years building under sustained regulatory pressure.
Kalshi’s $1 billion March Madness bracket offer – the one where the CEO used the phrase “generational wealth” in a 90-minute countdown – is not a financial product. It is a gambling promotion. The More Perfect Union video makes the observation directly: “If you’re betting on the Super Bowl coin toss, you aren’t investing. You’re gambling.”
What the licensed operator was told
I want to be precise about what happened to the gambling industry over the roughly fifteen years that produced its current compliance architecture, because the contrast matters.
Regulators made a specific argument. The gambling product creates structural asymmetry between operator and player. The player does not fully understand the probabilities. The marketing creates distorted expectations about outcomes. Certain customers, particularly those in financial difficulty, are vulnerable to harm that the product design does not naturally prevent. Therefore: affordability checks, safer gambling tools, self-exclusion registers, AML obligations, advertising standards, audit requirements.
None of this was painless. Platform selection became a compliance decision. Marketing budgets faced constraints that would have seemed absurd fifteen years earlier. Operators who had built their customer acquisition around particular promotional mechanics had to rebuild. Some of the regulatory architecture was badly designed and overcorrected in ways that damaged the customer experience without protecting the customers who most needed protection. The argument was adversarial and expensive and is still ongoing.
But the underlying premise – that a product creating structural asymmetry between sophisticated operator and retail customer requires protection for that customer – was not wrong. The licensed industry accepted it, built the architecture around it, and operates within it.
Now look at what More Perfect Union found. A product where 70% of retail users lose money. Where 0.04% of traders – the professional algorithmic firms – capture 70% of the profits. Where marketing explicitly targets working people with language about generational wealth and making rent. Where bets on government action are structured in a way that makes insider trading endemic and nearly impossible to prosecute.
The asymmetry is larger. The targeting is more explicit. The insider trading problem has no serious enforcement architecture. And the classification as “event contracts” rather than gambling means none of the player protection requirements the licensed industry was required to build apply here.
The question the industry is not asking publicly
There is an argument the licensed gambling industry should be making, and is mostly not making, because the political dynamics are uncomfortable.
The argument is this: we were required to build player protection infrastructure because our product created structural asymmetry with retail customers. That obligation cost real money and constrained our commercial model at the edges. We accepted it because the regulatory case for it was coherent.
Prediction markets are running a product with greater structural asymmetry, against the same retail customers, with more predatory marketing, with an insider trading problem that has no serious enforcement response. They are classified in a way that exempts them from the obligations we were required to build. The basis for the exemption – that these are financial instruments, not gambling – is contradicted by their own marketing.
The licensed operator who sits quietly while this plays out is not being diplomatic. They are subsidising a competitor’s regulatory arbitrage with their own compliance costs.
More Perfect Union published the video because Polymarket approached them for a partnership and they looked at the business model before saying yes. The 3 million people who have watched it did not need a regulator to tell them what the data showed. They needed a nonprofit newsroom with a YouTube channel and the willingness to read a spreadsheet.
That gap – between what the data shows and what the regulator has so far required – is the argument the licensed industry should be making. Loudly, and in writing, before someone else makes it for them.
