In late May, federal authorities charged a Google software engineer with insider trading after he won $1.2 million on the prediction-market website Polymarket. The 36-year-old Michele Spagnuolo allegedly placed bets that musician D4vd and rapper Kendrick Lamar would top Google’s most-searched list. The bets paid off, prosecutors said, because Spagnuolo had access to confidential company data. 

The popularity of prediction markets, where you can bet on thousands of real-world outcomes across nearly every facet of modern life, is spreading faster than governments can keep up. Even Mark Zuckerberg, Meta’s chief executive, is reportedly developing a standalone prediction market app to compete with the most popular platforms, Kalshi and Polymarket. 

You may have even been tempted yourself to put down cash on your favorite pop-culture hunch. But the recent Google case highlights just one of the biggest concerns for a multibillion-dollar industry prone to abuse. Numerous insider trading cases have prompted federal regulators to intensify scrutiny, cracking down on the illegal use of classified information for betting. 

A New York Times investigation in May flagged more than 11,000 Polymarket accounts for suspicious, high-profit trading patterns, often involving perfectly timed bets on geopolitical events, and flawless, loss-free track records. And it’s not just corporate employees; it’s also military personnel and government officials manipulating classified information. 

A US Army special forces soldier allegedly received a payout of $400,000 by “predicting” the capture of Venezuelan president Nicolas Maduro. Former Congressman George Santos allegedly won tens of thousands of dollars by betting he wouldn’t be at Trump’s State of the Union address, despite posting on X that he would be.

Just last week, a Wall Street Journal investigation revealed that Polymarket ran a deceptive, secret marketing campaign by paying social media influencers to film fake trades and stage massive winnings on lookalike dummy websites to draw people in.

“This industry is growing fast and will continue to grow as long as courts and regulators allow it,” Columbia University professor of economics Rajiv Sethi told CNET.

People generally have strong opinions surrounding prediction markets, and many (like me) feel a bit icky about them. But how the industry shakes out will depend on several regulatory battlegrounds. Prediction markets are facing intense pushback from lawmakers over insider trading, highlighted by a congressional probe and a proposed bill to ban prediction-market bets by service members. Yet because no one can agree whether betting markets are legitimate financial tools or just a glorified form of gambling, they’re causing a massive headache at the state and federal levels. 

How prediction markets work

To any casual observer, Polymarket and Kalshi seem like virtual casinos, except you’re betting against other participants, not against “the house.” You can buy and sell contracts about anything: the weather, geopolitical events, election results, sports, entertainment awards, ad nauseam. 

Several high-profile predictions over the past several months involved the US attacking Iran, Michael B. Jordan winning the Oscar for Best Actor and bitcoin topping $125,000. You can even predict if someone is going to utter a certain word in a speech or news conference in what are called “mention markets.” 

With a mainstream boom in prediction market platforms over the last few years, other companies have joined the fray: Robinhood, PredictIt, Metaculus and even traditional sportsbooks FanDuel and DraftKings. 

These types of “idea futures” aren’t new, though. Informal information markets date back hundreds of years, as seen in the 1500s in Italy, where people predicted who the next pope would be.

Today’s prediction markets claim they aren’t technically gambling or akin to trading stocks, even though you’re risking money in hopes of a profit. In essence, you’re predicting something will or won’t happen. For every “share” you buy for that event outcome, you get $1 if you’re right and nothing if you aren’t. The markets don’t set the “odds,” and neither do the platforms — the traders do.

The amount of shares you’re able to buy for a certain outcome depends on how many shares are being sold for the opposite outcome by other traders. For example, if you wanted to buy 500 shares of a Yes outcome on France winning the World Cup, there would have to be 500 corresponding shares of No on France winning. 

Though the basic unit for prediction markets is only $1, business is booming for Kalshi and Polymarket, which collect transaction fees for each trade. Together, they’ve crossed $150 billion in lifetime trading volume.

A personal look inside

I’m not a bettor. I suck at poker, I still can’t understand a Daily Racing Form, and don’t get me started about March Madness brackets. So, I’m not about to test my luck (yet) with Kalshi or Polymarket, but I did want to take a peek under the hood.

Kalshi and most other prediction markets are available for customers in all 50 states, but Polymarket, a cryptocurrency-based company, is restricted in the US and several other countries, at least for now. Some people try to bypass geographic restrictions using a VPN, even though Polymarket says it blocks VPN IP addresses.

Kalshi and Polymarket both offer a dizzying array of exchanges. Kalshi has basic event categories, from the California governor race to the price of a gallon of gas. It also has some rather off-the-wall ones, like the “Scary Tomatoes” score on Rotten Tomatoes and the US government’s disclosure of aliens. 

The mention markets on Kalshi were even stranger. Will someone say “road trip” or “meals for two” during the next Cracker Barrel earnings call? What will the hosts say during Love Island Aftersun? I’ll pass.

Columbia professor Sethi advises anyone interested in trading prediction markets to tread lightly at first.

“Most novice retail traders lose money, so my advice to those who want to experiment is to focus on events about which you know something about the topic, and keep bets small to begin with, until you get a feel for your likely performance,” Sethi told CNET.

The hard truth is that prediction market traders are far more likely to lose than to win. The Wall Street Journal reported in May that 0.1% of all Polymarket accounts won 67% of the profits. That translates to 2,000 top traders netting more than $500 million, while 1.1 million Polymarket customers didn’t make a profit.

Moreover, given the difficult task of preventing insider trading, I’d say “buyer beware” when trading in these speculative markets. 

Social function or political tool

Another fundamental question I have is whether these markets serve a socially useful purpose. 

Better Markets, a nonprofit focused on financial and economic justice, argues that prediction markets lack real value. While traditional financial contracts help institutions manage risks, prediction markets do not. Unlike the stock market, they fail to fund businesses or help investors build long-term wealth.

Amanda Fischer, chief operating officer at Better Markets, said that bets around elections or war in Iran “serve no function but to degrade our democracy and encourage insider trading.” According to Fischer, prediction markets look more like gambling, especially since over 90% of bets on those platforms are related to sporting events. 

In response to scandals around insider trading, Kalshi says it is aggressively self-policing by tracking suspicious activity and requiring some of its users to disclose their employers. Kalshi also says its safeguards against politicians and athletes are stricter than those of traditional stock exchanges.

Meanwhile, Polymarket’s decision to maintain user anonymity has drawn heavy criticism from financial experts, who argue it leaves the platform vulnerable to fraud. Without strict identity verification, the platform allows insiders to exploit nonpublic information while enabling bad actors to “spoof” trades and trick ordinary people into following fake trends, according to Sethi, who wrote an opinion piece for the Financial Times titled “Polymarket Anonymity Must End.”

As prediction markets continue to face security concerns over fraud and insider trading, they have a powerful shield from the federal government and President Trump, who has aggressively pushed back against state-level restrictions. 

This political alignment is further complicated by the president’s son, Donald Trump Jr., who reportedly has an eight-figure investment in Polymarket and serves as an adviser to Kalshi. Although his involvement has sparked intense suspicion of a conflict of interest, Trump Jr. maintains that he does not trade on the platforms or lobby the government on their behalf.

A regulatory dilemma 

At its core, the regulatory mess stems from an identity crisis. Prediction markets are hard to classify, straddling the line between commodity contracts and security-based investments. This has triggered a massive turf war over jurisdiction, as the federal government attempts to override state and tribal gaming laws that view these markets as illegal sportsbooks trying to bypass local restrictions.

Several US states and even private citizens have sued Kalshi, claiming the company has violated state gambling laws. Native American pueblos and a tribe in New Mexico have also sued Kalshi, alleging the company is violating gaming agreements and federal law.

Sandia Pueblo Gov. Stuart Paisano, one of the plaintiffs, in a statement, said, “The use of prediction markets for gambling purposes diverts essential revenue away from our governments, provides an end-run around regulation of gaming on our lands, and allows gaming by underage people.”

At the federal level, prediction markets are formally categorized as commodities and derivatives, placing them under the jurisdiction of the Commodity Futures Trading Commission, or CFTC. 

Kalshi CEO Tarek Mansour, who, along with fellow MIT graduate Luana Lopes, founded the company in 2018, says prediction markets aren’t traditional sportsbooks but more like open marketplaces. Mansour says Kalshi’s event contracts are financial derivatives, just like common futures, options and swaps, and should be appropriately regulated by the CFTC.

But some legal scholars and financial reform advocates argue that prediction markets should fall under the purview of the Securities and Exchange Commission, or SEC.

According to Better Markets’ Fischer, the CFTC has fewer tools to police insider trading in prediction markets. As an agency tasked with specifically overseeing agricultural and certain financial derivatives, it was only recently self-appointed as a gambling regulator. “As a result, there are some gaps and ambiguity in the CFTC’s legal framework,” she said. 

Fundamentally, the CFTC’s rules on insider trading are historically much weaker than the SEC’s. “The SEC has 90 years of law and legal precedent, which have created a robust set of rules around insider trading,” said Fischer. 

The CFTC is also chronically understaffed, according to Fischer. The agency has cut more than 20% of its staff during the second Trump administration.

Fischer said CFTC’s enforcement is a “drop in the bucket” compared with the enormous volume of trades being transacted at Kalshi. “The CFTC has only been able to identify and prosecute the most egregious cases, and in many other instances, has delegated enforcement to firms like Kalshi, whose only tool is to kick users off the platform,” Fischer said.

Do we really need this?

The danger of prediction markets is the financialization of our society at large, where “every opinion is a tradeable asset,” wrote Jathan Sadowski, associate professor at Monash University in Melbourne, Australia.

There’s also a risk if prediction markets define “truth” as simply a publicly verifiable consensus. If, as Sadowski noted, “the market is the ultimate arbiter of what’s valuable and true,” that leads to a “world that creates endless incentives for arbitrage, manipulation, collusion and exploitation in the pursuit of profit extraction.”

In an episode of Last Week Tonight on prediction markets, comedian John Oliver asked if we’ll be able to believe our eyes when future events occur. “When something unexpected happens in the world, it would be really nice not to have to automatically question whether it’s only because someone is trying to move a market.”

At the end of the day, I keep coming back to why these tools exist in the first place. Prediction markets shouldn’t just be a playground for day traders looking for their next fix. But to prove that it’s not just another corrupt form of speculative gambling, the industry has some massive hurdles to clear. 

CNET’s Laura Michelle Davis heavily contributed to and edited this story.