A week before the 2024 U.S. presidential election, traders on Kalshi moved nearly $2 billion in contracts betting on who would be Trump’s next Attorney General. On Polymarket, the same week saw almost as much volume swirling around Senate race outcomes and ballot initiatives. Meanwhile, Robinhood quietly rolled out its own prediction market hub—covering sports, crypto, climate, and even celebrity streaming numbers—with the same sleek interface users expect for stock trades.
What started as an academic curiosity—designers and economists theorizing about the "wisdom of the crowd"—has exploded into a real-time, $24 billion-a-month industry. And unlike traditional gambling or even sports betting, these platforms don’t make you visit a casino or open an account on a regulated exchange. You log into your phone, scroll through elections, sports scores, and crypto price ranges, and place a bet in seconds. The contracts look like investments: you buy "Yes" shares at $0.65 and wait for the event to resolve at $1—or lose your $0.65 if it doesn’t happen.
To early observers, prediction markets promised objective forecasting: financial stakes would force participants to research carefully, surface hidden information, and converge on accurate probabilities. The price of a contract was the crowd’s best guess about reality.
But today, as platforms shift from niche research tools to mass entertainment and financial Instruments, a growing chorus of counselors, regulators, and democratic integrity advocates warn that the same mechanics making prediction markets eerily accurate are also making them dangerously habit-forming. In a recent Policy Forum for Science, Nizan Packin and Sharon Rabinovitz called the current explosion a "large uncontrolled experiment," with behavioral risks that parallel slots machines—continuous play, variable ratio rewards, push notifications—and political risks ranging from foreign interference to insider trading on impending legislation.
This article walks through how prediction markets work, who’s building them, why volume exploded in 2025–2026, and what makes many experts think this is less a financial innovation than an urgent public health and democratic integrity moment waiting to escalate.
Mechanics: Binary Contracts, Real-Time Odds, and the $1 Payoff
At their core, prediction markets are just event derivatives—binary options tied to verifiable real-world outcomes. Users buy and sell shares representing whether an event will happen ("Yes") or not happen ("No"). The share price always sits between $0 and $1, reflecting the market’s collective assessment of probability. If a contract trades at $0.72, that means traders believe there’s roughly a 72% chance the event will occur.
Here’s how it plays out in practice:
- You buy a “Yes” share at $0.65 for a contract stating, “Will Team A win the 2026 NBA Finals?”
- If Team A wins, your share resolves at $1.00 and nets you a $0.35 profit per contract.
- If Team A loses, the share resolves at $0.00 and you lose your entire $0.65.
- If a breaking injury headline drops the next morning, the share price may slide to $0.52 overnight; you can sell your share at that lower price and lock in a $0.13 loss—or hold through resolution.
Unlike traditional betting, where you place a one-time wager and wait for the outcome, prediction markets let you trade your position continuously until resolution. The constant price shifts mimic stock trading, reinforcing the idea that participants are "investing" rather than “betting”—a distinction platforms emphasize to skirt stricter gambling regulations.
The behavioral parallels are deliberate, and increasingly documented. The Psychology Today piece by Amanda Giordano notes that the DSM-5 criteria for gambling disorder map neatly onto predictable patterns in prediction market usage: chasing losses after a losing streak, lying to conceal how much time or money is spent, and risking relationships because the market is “always open.”
It’s also worth highlighting what prediction markets are not. They aren’t pools where you bet against the house. You’re trading against other users, with the platform acting as an exchange—like Robinhood for stocks. That structural distinction has been central to platform marketing and regulatory avoidance, even as trading volume surges.
Platforms: Kalshi, Polymarket, and Robinhood’s Betting Hub
Three platforms dominate the current prediction market landscape—each with its own regulatory footprint, user base, and product design.
Kalshi launched in 2018 as a U.S.-regulated exchange overseen by the Commodity Futures Trading Commission (CFTC). It leans heavily into sports, with roughly 80% of its volume coming from leagues like the NBA, NFL, and college basketball. Politics accounts for just 4% of volume—partly because Kalshi waited until late 2024 to introduce political contracts around the presidential election. Crypto and broader economic indicators round out the rest.
A notable twist: Kalshi sued the CFTC in 2025 and won a preliminary injunction allowing it to continue operating, marking one of the first major legal challenges to how the CFTC has historically interpreted event-based contracts. This set off a wave of lobbying and internal review, culminating in the White House’s 2026 statement that it was reviewing CFTC proposals for broader regulation.
Polymarket, founded in 2020, is the leading Web3-native platform. It operates on-chain using smart contracts and, until recently, was completely unregulated in the U.S. (American users could—and still can—access it via VPN). In 2026, Polymarket acquired a CFTC-licensed exchange and launched Polymarket US, though that domestic version still only accounts for about 10% of the platform’s total volume ($1.3B in April 2026 vs. $9B on Polymarket International).
Polymarket’s topic mix is more diverse: sports (~39%), politics (~32%), and crypto (~20%). Its UI leans into the “decentralized finance” aesthetic, with token balances, gas fees, and on-chain resolution timestamps. It also recently partnered with Chainlink to power its oracle infrastructure—ensuring real-world outcomes feed accurately into on-chain smart contracts before settlement.
Robinhood’s Hub, launched in early 2026, brought prediction markets to hundreds of thousands of existing users overnight. The platform features 15+ categories, from esports and golf to presidential cabinets and Bitcoin price ranges. By integrating the hub into its main app, Robinhood blurred the line between stock trading and event forecasting—even including terms like “investing” and “buying shares” in its UI to reinforce the legitimacy angle.
According to Robinhood’s own page, it explicitly distinguishes prediction markets from traditional gambling by emphasizing peer-to-peer exchange and continuous trading. Still, the platform’s mobile-first design—push notifications, swiping to place orders, instant order confirmation—borrows heavily from mobile gambling UX studies that link engagement to addiction risk.
Together, these platforms have created a convergence point where finance, gaming, and political forecasting intersect—often with little clarity about which regulatory framework should apply.
Explosive Growth: From $5B to $24B in Less Than a Year
Pew Research Center’s analysis of data from The Block shows combined monthly global volume on Kalshi and Polymarket jumped from under $5 billion in September 2025 to about $24 billion in April 2026. That’s a four-fold increase in roughly seven months.
For context, legal sports betting in the U.S. averaged about $14 billion per month in 2025. So prediction markets are now outpacing legal sportsbooks by volume—even though they operate in a more legally ambiguous space, with fewer consumer protections and much lower transparency.
The acceleration aligns closely with a key 2024 U.S. court decision that allowed political event contracts on commercial platforms. Before then, prediction markets were largely relegated to university labs and private institutional tools. After the decision, platforms relaxed eligibility rules, dropped Know-Your-Customer barriers for international users, and ramped up marketing—especially around high-stakes events like the presidential election.
Notably, Polymarket’s international volume dwarfed its U.S.-regulated counterpart. In April 2026, $9 billion traded on Polymarket International compared with just $1.3 billion on Polymarket US. That asymmetry creates a regulatory arbitrage problem: American users gain exposure to high-volume, lightly regulated markets without the same CFTC oversight they’d expect from a domestic exchange.
What’s worrying to observers is not just the speed of growth, but what the growth is trading on. Events with thin participation—like state-level elections in small jurisdictions or niche crypto price bands—are vulnerable to manipulation. A single large trade can swing the market probability dramatically, creating false consensus signals that influence media narratives, campaign strategy, and even voter turnout in later stages. The same mechanisms making prediction markets accurate in liquid markets also make them dangerously malleable in thin ones.
Topic Distribution: Sports Dominates Kalshi, Politics Powers Polymarket
The topic mix varies significantly across platforms—and within platforms over time. Since July 2024:
- On Kalshi, sports account for ~80% of volume, crypto ~7%, and politics just ~4%. The dominance of sports tracks closely with real-world seasons: basketball and football surge during their respective leagues’ campaigns, while March Madness alone generated over $3 million in daily volume at its peak.
- On Polymarket, the distribution is more balanced: sports ~39%, politics ~32%, crypto ~20%. Political events (presidential debates, Senate runoff results, ballot initiatives) regularly drive 5–10% daily volume spikes.
Both platforms also host hundreds of niche topics: weather in major cities, celebrity mentions on social media, the number of tweets a figure issues in a day, even the exact date a company will hit a particular market cap. Most of these contribute only minor volume but demonstrate the ambition to tokenize any verifiable future event.
A telling pattern emerged during the 2024 U.S. election cycle: for a brief window in October and November, politics accounted for over 90% of Kalshi’s volume—simply because Kalshi hadn’t yet rolled out sports contracts. Once the election concluded, volume dropped sharply and shifted back toward other categories. That volatility hints at how dependent growth is on external events rather than steady, ongoing user demand.
Still, the emergence of themed markets—like “Climate” or “Metals” on Robinhood’s hub—suggests platforms are trying to build consistent engagement beyond headline events. If successful, that could stabilize volume and push prediction markets from_event-driven novelty to everyday financial behavior.
Gambling vs. Investing: Why the Distinction Is Blurring
The American Gaming Association (AGA) issued two reports in early 2026 that capture the tension. In one, they declare prediction markets are "offering illegal sports betting nationwide." In another, they cite internal survey data showing 28% of people who trade event contracts consider themselves "investing" rather than "betting."
That discrepancy isn’t just semantics—it’s how platforms avoid stricter gambling laws. Regulatory definitions vary, but the American Psychiatric Association’s DSM-5 defines gambling as "risking something of value in the hopes of obtaining something of greater value." By that definition, buying a $0.65 “Yes” share on Kalshi and hoping the event resolves in your favor is gambling.
Platforms counter by emphasizing three differences:
- Exchange vs. House: You’re not betting against the platform; it doesn’t take a side or guarantee payouts. Robinhood, for example, touts its peer-to-peer exchange model and ongoing trading before resolution.
- Liquidity and Pricing: Continuous price updates create a sense of financial activity—not just winning or losing at the end.
- Language: Phrases like "investing," "buying shares," and "position sizing" replace traditional gambling terms like "wagering" or "betting."
Yet research is catching up to the concern. In Science (2026), Packin and Rabinovitz describe how platforms act as “regulatory entrepreneurs,” designing features to straddle legal lines. Push notifications, variable rewards (e.g., a large payout on a long shot), and the illusion of control (“I’m just hedging my crypto portfolio”) mimic behavioral hooks already well-documented in gambling addiction literature.
Giordano’s Psychology Today piece spells out the clinical overlap: users meeting four or more of the DSM-5 gambling disorder criteria over a 12-month period show patterns identical to those seen in casino gamblers—chasing losses, lying about frequency, jeopardizing relationships. Counseling tools like motivational interviewing and CBT remain effective, but only if practitioners recognize the risk in a prediction market tab on someone’s phone.
What makes this especially urgent is that platforms don’t segment high-risk users. There are no "gambler" labels, cooling-off periods, or deposit limits in most prediction market apps. The default is open access, maximum engagement—exactly the recipe that sparked earlier gambling epidemics.
Democratic Integrity Risks: Foreign Interference, Thin Markets, and Insider Leverage
The democratic integrity risks fall into three interlocking categories: foreign influence, signal distortion, and insider advantage.
Foreign actors can legally trade on U.S. elections. Polymarket International is not subject to U.S. jurisdiction, and American users can access it with a VPN. There are currently no technical barriers preventing foreign individuals or state-aligned actors from buying contracts tied to election outcomes, Supreme Court rulings, or upcoming legislation. The CFTC’s 2026 review specifically cited election markets as a high-priority area for oversight.
Thinly traded markets amplify manipulation. On Kalshi’s political category, where volume hovers around 4%, a single $500,000 trade can shift the probability signal by several percentage points. That’s not just a market inefficiency—it becomes a media story. Headlines like “Poll Shows Biden Trailing After $500k Trump Trade” can influence real voters who see the market price as a leading indicator. The Science Policy Forum calls this "distorted consensus signaling," and it’s especially dangerous close to an election.
Insider trading gains new pathways. Because prediction markets settle on real-world outcomes, knowledge of pending government actions—like a Supreme Court decision, an upcoming regulatory filing, or internal agency memo—can be monetized instantly. Unlike traditional stock markets, where insider trading requires securities violations, prediction markets trade on any event-based contract. A trader who learns that a Senator is about to endorse a candidate can buy the relevant market share without violating securities law. The gap here is so large that regulators have begun calling it "event insider trading"—a category that doesn’t yet exist in statute.
Packin and Rabinovitz note that the current regulatory vacuum turns prediction markets into a "natural laboratory for testing behavioral manipulation." If platforms are designed to maximize user engagement—and engagement directly correlates with trading volume—the incentives line up against safeguards. The result is a dangerous feedback loop: as platforms improve their UX and marketing, they draw more users—some vulnerable to addiction or manipulation—which in turn increases volume and profitability.
The White House’s 2026 review of CFTC proposals is the first top-down acknowledgment that prediction markets may need their own regulatory category, distinct from sports betting or traditional financial derivatives. Until then, the platforms operate in a grey zone where finance, gambling, and public discourse all collide.
Behavioral Risks and Counseling Implications
The behavioral risks aren’t hypothetical—they’re measurable, and they’re accelerating.
Louis et al. (2026) tracked sports betting ads across digital platforms and found a direct correlation between ad exposure and self-reported urge to bet, especially among youth and people with pre-existing addiction or mental health conditions. Prediction markets amplify this effect by making those ads indistinguishable from financial education content. A Robinhood user scrolling through crypto predictions may not realize they’re being served a “gambling-adjacent” experience.
Packin and Rabinovitz note that counselors often miss these signals because prediction markets don’t look like traditional gambling. No casino chips, no slot machines—just a phone interface that mirrors trading stocks. This makes it easier for users to rationalize behavior (“I’m just hedging my portfolio”) and harder for clinicians to diagnose early-stage problems.
Giordano outlines four behavioral addiction hallmarks counselors should watch for in any client who trades event contracts:
- Compulsivity—accessing the app outside intended times, inability to stop despite intent
- Loss of control—increasing trade sizes or frequency to get the same emotional lift
- Continued despite negative consequences—arguing over money, missing work or school to trade
- Craving/mental preoccupation—spending free time analyzing odds instead of engaging socially
Tools like motivational interviewing remain effective, but counselors need concrete references: a client might not recognize “chasing losses” in crypto prediction contracts, even if they’d admit to it at a roulette table. The normalization of “investing” language makes intervention more delicate—and more necessary.
The public health community is responding. EurekAlert’s summary of Packin and Rabinovitz’s Science paper quotes a researcher warning: “Regulatory failure is structural—the stigma around behavioral addictions means funding lags far behind need, while powerful financial interests shape the evidence base to favor permissiveness. Without greater public investment in independent research, prediction markets will repeat the same pattern: innovate, normalize, harm—then finally regulate.”
What’s unique to prediction markets is the 24/7 availability and constant signal updates. Unlike a casino that closes at 2 AM, these platforms run in real time alongside major news cycles. A breaking headline can trigger a $0.20 swing in a market, prompting impulsive trades. That real-time urgency mimics the variable ratio reinforcement that studies show is most addictive of all.
Regulatory Outlook: The Looming Decision Point
As of mid-2026, regulators stand at a fork. The CFTC is reviewing proposals for broader oversight of event-based contracts, including position limits, residency checks, and marketing restrictions. The White House has publicly endorsed a “precautionary approach,” citing the Science Policy Forum.
Three regulatory paths are under discussion:
- ** classify prediction markets as gambling**, subjecting them to existing state laws and licensing requirements;
- create a new category—“event derivatives”— with bespoke capital reserves, age restrictions, and real-time volume caps;
- lean into the forecasting utility by licensing certain platforms as "official aggregators" for government or academic use, while banning consumer access to political and high-risk categories.
The third option reflects the original vision of prediction markets as tools for collective intelligence—not entertainment. But it’s unclear whether platforms like Polymarket and Kalshi would voluntarily cede their mass-market appeal for a niche, institutional role.
In the meantime, the data keeps climbing. April 2026’s $24 billion in combined volume is already more than double what traditional sportsbooks handled that month. If growth continues at the same pace, volume could top $100 billion annually within two years.
That scale—on an unregulated, globally accessible platform—makes the regulatory pause untenable. Not because prediction markets are inherently dangerous, but because their current design prioritizes engagement over well-being, liquidity over integrity, and profit over precaution.
The coming months will be pivotal. Either platforms and regulators collaborate to introduce guardrails—age verification, loss limits, transparency on thin markets—or the current trajectory sets off a broader reckoning about how society should treat behaviors that straddle finance, gaming, and public discourse.
Until then, counselors, policymakers, and ordinary users remain in the middle: watching a large experiment unfold, not knowing whether they’re witnessing a financial revolution or a preventable public health crisis.
Conclusion: Precision forecasting—or precision exploitation?
Prediction markets work astonishingly well. They aggregate information, reflect real-time probability shifts, and often outperform expert polling. That’s not in doubt.
What is in doubt is whether the platforms built around them should operate under financial, gambling, or no regulations at all. The platforms themselves have an interest inblurring boundaries—"investing" feels safe, "trading" sounds smart, and "betting" invites scrutiny.
The evidence suggests we’re in a transition zone: from niche forecasting tools to mass entertainment with gambling mechanics, all packaged as financial innovation. The rise of platforms like Robinhood and the CFTC’s growing scrutiny suggest that transition is nearing its endpoint—either regulation catches up, or the problems accelerate.
For now, the best defense is awareness: understanding that buying a $0.65 "Yes" share on a binary contract is gambling by definition, even if the UI says otherwise. Clinicians who recognize the DSM-5 criteria can intervene earlier. Regulators who separate signal from manipulation can protect democratic integrity. And users who know how the mechanics work can make informed choices about whether they’re forecasting—or being forecasted.