How the Federal Rules for Prediction Markets Fall Short of State-Level Gambling Laws

September 22, 2026

A Roosevelt Institute banner reads: How the Federal Rules for Prediction Markets Fall Short of State-Level Gambling Laws on a dark grid background with colorful digital graphics on the right.

Prediction markets offering contracts on virtually every topic are one of the fastest-growing industries in the United States, just trailing behind AI. Who wins and loses in these markets? How is their operation shaping our society? And what rules do we need to manage them?

In a new four-part series, The Hidden House: Prediction Markets and How They’re Shaping Society, we’re exploring all of these questions and more.

In this fourth installment, we compare the rules that apply to prediction markets to those that apply to traditional gambling.

When casual users bet on everyday events on prediction market platforms, that’s a form of gambling. But although the platforms have a lot in common with casinos and other types of traditional gambling, the Trump administration and the platforms themselves are trying to establish a very different set of rules. Locked in a pitched battle with state regulators across the country, prediction markets are aiming to displace state law in favor of a dramatically different, federal-only standard.

The rules around traditional gambling leave a lot to be desired, especially for preventing “problem gambling” and addiction. Nonetheless, prediction markets seek to be governed by even more limited rules, giving people exceptionally few safeguards when they participate in this form of gambling.

Despite the legal warfare between state regulators and prediction market platforms, we have not seen an analysis of the disparities between state laws and the emerging federal regime. How different will consumer protections be if prediction markets only have to follow the federal rules?

So that policymakers and the public can see the difference, this post examines the restrictions that states like Nevada and New Jersey apply to traditional gambling in places like Las Vegas and Atlantic City and how they compare to the activities that take place on prediction markets—which argue that they only must comply with the federal rules for “swaps.”

In particular, prediction markets’ ability under the federal rules to offer bets on an incredible range of topics, as well as the lack of any disclosure about people’s chances of winning or who they are betting against, stand out as areas where people have decidedly weaker protections than they would under the rules of traditional gambling.

Prediction Markets Don’t Want to Be Subject to State Gambling Laws

Gambling is regulated by the states, so traditional gambling outlets such as casinos accept that they are governed by state law.

In contrast, although much of the activity on prediction markets is basically just gambling by nonprofessionals, Kalshi, the largest US prediction market platform, contends that it is offering swaps—financial instruments used by sophisticated investors, often to hedge risk—and that only the federal rules applying to swaps should apply to that activity, completely displacing state law.

A wide range of states, including Kentucky, Nevada, and New York, have challenged this view, suing or taking other legal action against prediction markets.

The Trump administration, in what seems to be an unprecedented move, has sued states that try to enforce their own laws against prediction market platforms. Courts have split on the question of whether state law does in fact apply, but most have ruled that it does, with New York, Utah, and Nevada winning recent skirmishes.

A Key Distinction Between Prediction Markets and Traditional Gambling: What People Can Bet On

Text about prediction market regulations overlaid on a grayscale background of a betting shop, with highlighted text and warning icons emphasizing regulatory and enforcement issues. Roosevelt Institute logo appears in the corner. Quote Reads: In contrast, under federal law, prediction markets are required to report to regulators what types of bets they are offering, but they can offer those bets without approval from regulators under the self-certification model. Contracts are theoretically prohibited if they relate to “terrorism, assassination, war, gaming,” or illegal activity, or are “readily susceptible to manipulation,” but it isn’t clear how meaningfully those restrictions are being enforced.

As long as prediction markets fail to comply with state law, one of the biggest differences between prediction market platforms and traditional gambling is what can be bet on.

At the state level, gambling operators must generally get permission before operating particular types of bets, whereas prediction markets are currently “self-certifying” bets to federal regulators. In practice, this means that instead of public deliberation about what types of gambling are appropriate and a closed universe of available bets, prediction market bets are available on almost any topic.

The Nevada Gaming Commission maintains a database of allowed gambling games, which includes specifications about the rules and odds of those games. New Jersey law similarly limits gambling to “authorized games” approved by state regulators. Las Vegas and Atlantic City, two of the largest gambling markets in the country, both generally allow sports gambling (subject to limitations, such as some restrictions on betting on amateur sports). But in Nevada, sportsbooks must get permission from regulators before offering bets on other types of events, while New Jersey strictly limits gambling to certain types of events. (Indeed, perhaps preferring the regulatory regime around prediction markets over what applies to traditional gambling, the two largest online gambling platforms, DraftKings and FanDuel, each recently announced they are giving up on getting licenses in Nevada and are opening prediction market platforms instead.) Many state laws also prohibit gambling on other topics, such as elections.

In contrast, under federal law, prediction markets are required to report to regulators what types of bets they are offering, but they can offer those bets without approval from regulators under the self-certification model. Contracts are theoretically prohibited if they relate to “terrorism, assassination, war, gaming,” or illegal activity, or are “readily susceptible to manipulation,” but it isn’t clear how meaningfully those restrictions are being enforced. (The Commodity Futures Trading Commission (CFTC) does not appear to have challenged a single self-certified bet on Kalshi or Polymarket, the two biggest prediction markets, since their use exploded in the past few years.)

As a result, there are now a dizzying array of prediction market bets available on almost any topic. As Roosevelt has noted, the sheer range of available bets raises some significant issues, including the possibility of financially incentivizing people to change their behavior on a wide range of activities. (The Trump administration is now considering some fairly limited additional restrictions on prediction markets contracts, such as prohibiting gambling on injuries to players.)

Insider trading and other forms of cheating are prohibited at both the state and federal level for all forms of gambling including prediction markets, as is deceptive advertising, though again it isn’t clear how well these rules are being enforced. Additionally, gambling is restricted to people over 21 by law in both Nevada and New Jersey, but prediction market platforms allow anyone age 18 and up to participate.

For Prediction Market Users, the Odds of Winning Are a Black Box

Casinos in Nevada and New Jersey are governed by detailed rules about what must be disclosed to consumers making bets. In Vegas, for example, “payoff schedules”—which document how much a bet pays off based on which outcome occurs—generally must be “either on the table or machine or in a conspicuous place immediately adjacent.” This means that a player could at least theoretically calculate the odds of winning a particular bet for table games.

The same is true for slot machines that digitally represent physical objects, such as cards or dice, because the probabilities of pulling a card or getting a number on dice must be the same on the machine as with the physical object. On the other hand, casinos have been criticized for making it difficult or even impossible to know the odds of winning in popular video slot games that don’t represent physical objects, as well as for manipulating “near misses” to make it look like players were closer to winning than they really were. This manipulation can encourage addictive gambling.

Notwithstanding these shortcomings of the rules governing traditional gambling, on prediction market platforms it is now basically impossible to tell what the odds are of winning a wager. Participants are only told what amount they will be paid if they win a particular bet. Although research from the Roosevelt Institute and elsewhere suggests that ordinary participants do poorly competing against professional traders on these platforms, there is no disclosure about the financial backing of people on the platform, how often they bet, or how often they win or lose.

Prediction Markets Lack Protections Against Addiction

Studies have estimated that as much as 60 percent of gambling revenue is from “problem gambling” by those who are addicted. In Las Vegas, casinos are required to post materials referencing the industry-funded Nevada Council on Problem Gambling and train staff on problem gambling behaviors. These rules have been criticized as woefully inadequate to address addictive behaviors. Nonetheless, there are even fewer rules—indeed, essentially none—with respect to addictive gambling on prediction markets, since those platforms are contending they need to comply only with the federal rules that apply to sophisticated financial instruments. 

A quote about Kalshi and problem gambling is shown in bold white text on a dark background with graphics of warning symbols and redacted screenshots. Logos of Roosevelt Institute and Fanatics are also visible. Quote reads: The issue of problem gambling reveals the shortcomings of Kalshi’s insistence that its product be regulated like a sophisticated financial instrument available to anyone who is 18 or older.

Moreover, the sheer prevalence of advertising for prediction markets—including content from paid influencers that may not immediately look like advertising—is raising concerns about drawing in or creating problem gamblers, including 18-year-olds who cannot legally participate in traditional gambling. (Kalshi has announced that it will voluntarily adopt some problem gambling measures, primarily focused on letting people “self-exclude” from using its platform.)

The issue of problem gambling reveals the shortcomings of Kalshi’s insistence that its product be regulated like a sophisticated financial instrument available to anyone who is 18 or older. (As we noted in another context, the regulator of those instruments, the CFTC, typically deals with a business-to-business market.)

What Policymakers Should Consider for Regulating Prediction Markets

There is significant reason for concern about prediction market activities and their broader influence on our society. Ordinary users on prediction markets are losing a lot of money to professional traders, without much disclosure about who they are really betting against or how often different users win or lose. 

Prediction markets are also giving people financial incentives that could affect a wide range of real-world events, from politics to matters of war. 

A sports bar with large screens, Roosevelt Institute logo, and overlaid graphics; text discusses applying state gambling rules to prediction markets while improving regulations. Quote reads: The most logical outcome may be to apply the state rules around traditional gambling to prediction markets while simultaneously improving those rules.

Further, as these platforms take off, there is a risk that they will become increasingly embedded in our society—and more difficult to regulate in the future. Policymakers should take a close look at what is happening on these platforms and consider appropriate reforms.

As discussed, the activities and functions of prediction markets diverge from the rules that states apply to traditional gambling. Our analysis suggests that while traditional state-level gambling regulations could be significantly improved, they are nonetheless more robust than the decidedly minimal regulations that prediction market platforms and the Trump administration are seeking to apply—the federal regime that applies to swaps. 

A lot of prediction market activity is quite similar to traditional gambling, and it’s unclear why different rules should apply. The most logical outcome may be to apply the state rules around traditional gambling to prediction markets while simultaneously improving those rules. That would dramatically curtail the range of bets available on prediction markets, and could also help implement more protections against problem gambling and addiction and provide platform users with more information about the odds of winning and who they are really betting against.

Related Resources