On Polymarket, 0.1% of accounts capture 67% of all profits [1]. Think about that for a second. While millions of ordinary people plowed money into election bets and sports wagers, a tiny slice of mega-players walked away with the lion's share. And the data suggests most of them were running AI-powered bots designed to extract value from everyone else. This is the uncomfortable reality behind the "democratization of prediction markets" narrative that platforms like Polymarket and Kalshi have been selling.

The prediction market boom is real. Polymarket reached a $1 billion weekly trading volume peak during the 2024 US presidential race, with over $3.3 billion wagered on the outcome between Donald Trump and Kamala Harris [1][3]. Kalshi, its domestic competitor, hit a $22 billion valuation in May 2026 following a TCV-led funding round [2]. By some measures, roughly $30 billion in notional value moved through these platforms recently [4]. But understanding who actually benefits from this growth matters more than the headline numbers.

The User Base Breakdown

Walk into any prediction market platform and you're unlikely to find a representative cross-section of America. Young male traders aged 18 to 34 dominate these platforms, comprising approximately 70% of all users [5]. That's a demographic profile that skews heavily toward risk tolerance, crypto familiarity, and free time to monitor markets. It's not the "everyone" that executives like to invoke when they talk about "democratizing forecasting."

Sports betting accounts for the vast majority of activity on both major platforms. On Polymarket, 63% of all trades involve sports bets [1]. On Kalshi, more than 90% of activity falls into this category, with 89% of the company's 2025 revenue coming from sports wagering specifically [2]. These aren't sophisticated information markets where crowd wisdom aggregates to produce accurate forecasts. They're essentially sportsbooks wearing a Silicon Valley aesthetic.

The companies themselves have made this clear. Kalshi's co-founder Tarek Mansour stated the long-term vision is "to financialize everything and create a tradeable asset out of any difference in opinion" [2]. That's a candid admission that the real product isn't prediction accuracy, it's transaction volume.

The Profit Distribution Problem

Here is where the story gets uncomfortable for retail participants. On Polymarket, more than 70% of users lose money [1]. The top 0.1% of accounts by profit concentration capture 67% of all gains [1]. An estimated 3% of traders account for the majority of price discovery on the platform [1]. This is not a crowd-sourced wisdom engine. It's a zero-sum market where sophisticated players systematically extract value from less sophisticated ones.

The mechanism isn't subtle. Hedge funds now deploy AI bots that scan social media and news to place trades on prediction markets automatically [4][5]. Regular retail traders are competing against algorithms with faster execution, better data processing, and more capital. A YouTube video editor was fined and suspended for suspected insider trading on Kalshi in February 2026 [2]. Three congressional candidates were fined in April 2026 for betting on their own races [2]. These aren't edge cases. They're symptoms of a market structure that rewards information advantages.

The Polymarket French trader who won $85 million when Trump won the 2024 election illustrates the scale of what's possible for well-capitalized participants [1]. Four accounts controlled by this single trader drove an odds spike that preceded Trump's actual victory. That's not crowd wisdom. That's someone with enough capital and sophistication to move prices and profit from the movement.

The Insider Trading Problem

Both platforms have struggled with insider trading in ways that should concern regular users. Military members have placed bets on operations, war reporting, and nuclear war using classified information [1]. The 52% success rate of military action long-shot bets valued at $2,500 or more with odds at or below 35% compares to just 25% for political bets and 14% for all bets [1]. This gap suggests some traders had access to information that ordinary participants could not possibly have.

The consequences have been real. On Kalshi, $77 million in winnings were frozen when Iran's Supreme Leader Ali Khamenei died [2]. A proposed class action lawsuit filed in November 2025 alleged Kalshi was operating unlicensed sports betting and engaging in deceptive practices [2]. Multiple states have taken action: Massachusetts obtained a preliminary injunction in January 2026, Wisconsin filed lawsuits in April 2026, Arizona filed criminal charges in March 2026, and Nevada issued a temporary ban also in March 2026 [2]. The US Senate banned its members and staff from using Kalshi in May 2026 [2].

Can Regular People Actually Profit?

The honest answer is: probably not, for most people. Prediction markets have been shown to outperform opinion polls in accuracy [5], which suggests they do aggregate information effectively. But that doesn't mean the average person can systematically profit from them.

The platforms are designed to extract fees on every trade. The information advantages held by hedge funds with AI bots are substantial. And the profit concentration data makes clear that a small number of players are taking most of the money out. For casual users treating prediction markets as entertainment or a way to add excitement to the election cycle, that's fine. For anyone treating it as an investment strategy, the data suggests they'll be feeding someone else's returns.

Regulatory Wild West

Both Polymarket and Kalshi have operated in regulatory grey areas that have now become more active enforcement targets. Polymarket was fined $1.4 million by the CFTC in January 2022 and blocked US customers until December 2025 [1]. It remains banned in France and Brazil [1]. Donald Trump Jr. joined as an advisor after his firm invested, and the Trump administration eased the regulatory environment [1].

For now, the sector is booming despite, or perhaps because of, its regulatory ambiguity. But the accumulation of state-level actions, congressional scrutiny, and insider trading cases suggests that the "wild west" period may be ending. Whether that makes these markets safer for ordinary participants or simply concentrates them further among sophisticated players remains to be seen.