Prediction markets sell themselves as aggregators of information, but their risk is that they can turn events into tradable temptation. The central problem is not only gambling, but market integrity under extreme information asymmetry.
Recent scrutiny of Polymarket and Kalshi after well-timed bets on Venezuela and Iran showed why critics are alarmed. When contracts hinge on military action, political upheaval, or policy surprises, the line between forecasting and profiting from privileged knowledge starts to blur fast. These platforms may look innovative, but the closer they move to real power, the harder it becomes to separate price discovery from ethical hazard.
Where the sharpest risks sit
Insider trading is the fault line. A market that rewards unique information can also reward information that should never be tradable. Reuters reported that unknown traders made more than $400,000 on a Venezuelan regime-change contract and that six accounts made $1.2 million from bets funded before the strikes that killed Iran’s supreme leader. The CFTC’s enforcement director has said insider trading in prediction markets is a priority, and California barred state officials from using nonpublic information to bet on these venues. That matters. A market cannot claim to reveal truth if participants suspect someone already knows the answer.
The second risk is simpler and more brutal. Most traders are not as informed or skilled as the market makes them feel. A 2025 study using more than 124 million Polymarket trades found that only 30% of traders earned positive profits, and that successful traders appeared to exploit the biases of less skilled participants. The same paper found prices tracked probabilities well, which is precisely why the platform is dangerous for beginners. Informative markets can still be costly markets. Users may mistake a probability number for a fair field, even when their counterparties are faster, smarter, and more disciplined today.
Why the next phase will be judged differently
Regulation is the third hazard because uncertainty changes behavior. These platforms sit in a jurisdictional gray zone just when their social consequences are getting larger. Reuters reported in March that the CFTC opened a rulemaking process asking how to regulate event contracts, including questions about manipulation, margin trading, terrorism, military action, and the role of insider information from government employees. At the same time, the agency is fighting states that want to treat some contracts more like gambling. That unresolved perimeter is not a technicality. It shapes surveillance standards, disclosure expectations, enforcement capacity, and public trust across rapidly growing markets.
My view is that decentralized prediction markets are underpriced as a governance problem. Their future depends less on popularity than on whether they can prove fairness under stress. The bullish case is obvious: these markets can aggregate scattered information better than pundits or polls. But once contracts move close to war, state action, or confidential policy decisions, the social cost of informed trading rises sharply. Prediction markets will not disappear. They may become infrastructure. Yet if platforms such as Polymarket or Kalshi cannot contain manipulation, opacity, and insider advantage, they risk becoming less like forecasting tools and more like casinos.