Prediction Markets: Last Week Tonight with John Oliver (HBO)
Prediction markets are booming platforms where people "trade" on the outcomes of real-world events—but John Oliver argues they often look and behave like gambling, including on war, tragedy, and death. He walks through where they came from, how companies like Kalshi and Polymarket claim legitimacy, and why their legal status is so messy (and politically convenient). Along the way, he highlights major risks: manipulation, insider trading, misleading media partnerships, and the moral cost of turning everything into a wager.
1. A fast-growing way to bet on basically anything 📈🎲
John opens by introducing prediction markets as "fast growing platforms where you can bet on basically anything," and immediately shows how casually people treat them—like just another app.
A TikToker lists a string of wildly varied bets: video games, elections, weather, celebrity trials, airline "Zen," and more. The point is not whether those bets are smart, but that the menu of wagerable reality has become huge.
"I've been making a ton of money on this app that lets you bet on anything in the world."
John jokes that the guy's interests are a perfect split between Gen Z boy (GTA, podcasts, nicotine pouches) and suburban mom (weather, temperature, celebrity court drama).
"Caden, you contain multitudes."
Then he defines the concept clearly: prediction markets let you wager on the outcome of future events, with companies choosing what questions to list. He mentions examples ranging from shipping traffic returning to normal to whether the U.S. will confirm aliens exist by 2027. He stresses how huge they've gotten—billions of dollars in bets every week—and how marketing plus influencer culture is pulling more people in.
We see an influencer loudly reacting to news because his bets depend on it, and John highlights how bizarre it is to watch someone emotionally melt down over a market outcome.
"A lot more screaming than you thought there'd be."
The segment turns darker when John notes you can bet on grim events—like leadership upheaval in Iran or whether a kidnapping suspect will be arrested by a certain date—making tragedy feel like content.
"Being able to bet on… horrible things like that… that's really dark to me."
John agrees, emphasizing the basic moral recoil: when someone dies, the appropriate response is sympathy, not celebrating your winnings.
"You're supposed to send their family a card that says, 'Sorry for your loss,' not one that says, 'Thanks for covering the spread.'"
He frames the rest of the story as three big questions: Where did these sites come from? What purpose do they serve (if any)? And how is this legal?
2. Where prediction markets came from (and why this isn't totally new) 🧠📊
John explains that the broad idea of betting on events to forecast outcomes is old. In the late 1980s, economists at the University of Iowa created the Iowa Electronic Markets, where people traded predictions about elections. Interestingly, those markets sometimes beat traditional polls—based on the "wisdom of crowds," especially when people put real money behind their beliefs.
He also notes that controversial versions existed long before today's apps. After 9/11, the Pentagon briefly considered a market where experts could bet on Middle East events—then got shut down fast once Congress heard what it would include.
A member of Congress reacts with disgust at examples like assassination or missile attacks being turned into wagers.
"This is unbelievably stupid… it combines the worst of all of our instincts."
John's punchline is that society has historically reserved "profiting from instability" for certain already-powerful groups, not random bettors—underscoring the hypocrisy in what we consider normal.
3. Kalshi vs. Polymarket: two paths into a booming industry 🚪💰
John explains how things shifted "a few years ago" when Kalshi and Polymarket launched and opened the door to today's free-for-all.
- Kalshi pursued regulatory approval and became (in 2020) the first regulated exchange in this space.
- Polymarket launched without approval and operated in a more anything-goes style.
Polymarket got hit by regulators for operating an unlicensed exchange, and in a 2022 settlement it paid $1.4 million and agreed to block Americans (while not admitting wrongdoing). John plays a clip of CEO Shayne Coplan trying to minimize it.
Coplan's defense is basically: it wasn't "breaking the law," it was "incompatible with the regulatory matrix"—and he even says, "Which law?"
"People say breaking the law. It's like, which law? You know."
John mocks how absurd that is: if you paid a penalty, it's pretty clear some law applied.
"I'm pretty confident there is a specific law you broke given… you paid a $1.4 million penalty."
Polymarket's model is described as more libertarian: anonymous crypto-based betting, offshore operations (Panama), and Americans needing to mask their location to use the international version—something John jokes is easy for young people via a VPN.
4. The 2024 election hype and the marketing arms race 📺🔥
Both platforms got a massive boost after the 2024 election, when markets appeared to "call" the race before TV networks. John notes Kalshi and Polymarket both predicted Trump would win, and Polymarket's CEO posted a grandiose claim.
"Make no mistake, Polymarket single-handedly called the election… The global truth machine is here, powered by the people."
John then pivots to how aggressively these companies marketed themselves afterward. He shows a chaotic Kalshi TV ad stuffed with bizarre imagery and people yelling predictions.
The ad ends with the tagline: "The world's gone mad. Trade it." John argues that, given the tone, a more honest message would be ruthless profit-seeking.
"A more honest tagline might have been, 'Fuck empathy, get money.'"
5. "This isn't gambling," they insist—so what is it? 🏛️📉
John addresses the core argument these companies make: they're not gambling sites.
They claim:
- Gambling is "against the house," but here users trade against each other.
- The platform just charges fees.
- These are like commodity futures markets (bets on future prices of corn, natural gas, etc.).
- Their "event contracts" let people hedge (protect themselves) against risks.
Kalshi's CEO gives an example: a market on whether student loans will be forgiven, where someone could hedge—if forgiveness doesn't happen, they win money that could help repay loans.
"A lot of people were hedging using that market."
John's response is basically: that logic might apply sometimes, but it collapses when markets are about celebrity nonsense or pure entertainment outcomes.
He also shows "customers in action," where people appear to bet on whether Trump will say certain words—treating speeches like a slot machine.
John calls it less "risk management" and more exploiting a person's declining coherence.
"That is not… hedging risk. It's taking advantage of… rapidly declining verbal abilities."
Then he points out a key reality: even if political markets get attention, about 90% of Kalshi's trading volume is sports, which makes it look even more like a sports betting business.
Why fight so hard to avoid the word "gambling"? Because gambling legality is patchy: many states allow some sports betting, fewer allow online betting, and huge states like California and Texas are lucrative targets. By calling themselves financial exchanges, prediction markets can sometimes operate where gambling is illegal—and even dodge state taxes and age requirements.
John highlights how slippery their messaging can be. Kalshi called itself the first nationwide legal sports betting platform, then later tried to walk it back by saying "bet" is just a metaphor for any financial position.
"That is so dumb it is almost charming."
His analogy: it's like a strip club claiming "girls, girls, girls" meant "girls who code" to support women in STEM.
6. The "we predict the future" claim—and why it's oversold 🔮📰
John says the two most irritating claims are:
- "We're not gambling."
- "We're vital to society because we predict the future better than anything else."
Polymarket's leadership claims it's "the most accurate thing we have" and even suggests people use it for safety decisions during conflict.
"We're looking at Polymarket to decide whether we sleep near the bomb shelter."
John is horrified at the idea of a billionaire CEO framing betting odds on bombings as a "value proposition."
"What exact stage of capitalism are we in… when… betting odds on bombings [are] an undeniable value proposition?"
He also mocks Coplan's styling and emphasizes a broader point: being surrounded by yes-men can make someone believe absurd things without anyone stopping them.
Then John brings nuance: prediction markets can sometimes match professional forecasters (a paper found they were roughly consistent on economic and Federal Reserve policy predictions), but they also miss badly—like Super Bowl outcomes or a Texas Republican Senate primary where Polymarket pushed an 83% projection that didn't happen.
The bigger issue is that companies market themselves as oracles, then amplify credibility through paid partnerships with news organizations:
- Dow Jones / Wall Street Journal partnered with Polymarket (earlier in the year—so, in 2026 context, that means 2026).
- CNN, CNBC, and Fox signed deals with Kalshi.
- CNN even ran Kalshi odds on a ticker.
John strongly rejects that idea: he doesn't want "betting odds on news events" presented like journalism.
"I DO NOT WANT THE CNN TICKER to have betting odds on news events."
He argues influencer promotions should be more transparent, then makes a joke by doing an "ad" disclosure himself—highlighting how easily audiences can be nudged.
Crucially, he notes that—just like sportsbooks—most users lose money. And winnings concentrate among a tiny fraction of accounts. One analysis found that though Polymarket had over two million users, more than two-thirds of all money won was held by just 740 accounts.
7. How politics and regulators keep the door open 🏛️🔓
John suggests the permissive environment "has a lot to do with who is in the White House right now," stating both companies have fostered strong connections with the Trump family.
- Donald Trump Jr. is an investor + unpaid adviser to Polymarket.
- He is also a paid adviser to Kalshi.
John then shows how evasive Kalshi's CEO is when asked what advice Don Jr. gives and what his role really is. The CEO keeps repeating variations of "we have a lot of advisers," and "it's about growing prediction markets."
"We have a lot of advisers."
John underlines why that's different: it's not "just any adviser"—it's a direct political pipeline.
"Here's a direct line to the White House."
He jokes that it's hard to justify hiring a nepo baby—"or in Don Jr.'s case, a nepo divorced dad"—but argues the bet paid off, because enforcement shifted dramatically between administrations.
The CFTC problem (Commodity Futures Trading Commission)
John explains the CFTC regulates these platforms. Under Dodd-Frank (2010), it was empowered to block event contracts about terrorism, assassination, war, and gaming—exactly the kinds of markets we're seeing.
But it's "clearly not doing that." Even worse, the CFTC is supposed to have five commissioners with bipartisan representation, yet currently it only has one: Michael Selig, a Trump appointee, presented as a major cheerleader for prediction markets.
John says Selig buys the "hedging" logic so hard he suggested people with medical conditions could hedge future treatment costs by betting on drug trial outcomes—John calls that one of the most disturbing proposals imaginable in the context of America's healthcare system.
Selig also promised during confirmation hearings he'd let courts decide whether these markets are gambling, but then he did the opposite—posting a video threatening challengers.
"To those who seek to challenge our authority… We will see you in court."
John calls him a weasel and notes the agency even sued states trying to regulate prediction markets—creating "extremely friendly conditions" for the companies.
8. The biggest danger: manipulation and insider trading (and it's already happening) 🧨💡
John moves into what happens when you let people bet on events they can influence—or events where someone has secret knowledge.
8.1. Manipulation can be ridiculously easy
He plays a clip where Coinbase's CEO jokes he was tracking a prediction market about what Coinbase would say on its earnings call—then intentionally says keywords to affect the outcome.
"I just want to… add here the words Bitcoin, Ethereum, blockchain, staking, and web 3…"
John points out: manipulating outcomes should be harder than just reading a word list.
"In the old days, you at least had to sneak cocaine to a racehorse."
He also reveals that bets were placed on what John Oliver himself would say on a specific show date—despite the fact the show is taped in front of a studio audience, giving them a huge information advantage before the betting window closed. (He jokes they "wipe" audience memories afterward.)
8.2. A real-world example: the WNBA dildo-throwing bets
John brings up a bizarre incident: people threw dildos onto WNBA courts to promote a crypto memecoin—and then Polymarket offered bets on whether a dildo would be thrown at specific games.
A clip shows someone explaining you can bet on that.
"You could place a bet on whether a dildo would be thrown…"
John emphasizes how cursed it is that someone learned the WNBA exists because they could bet on dildo-throwing. He adds Polymarket even promoted the bet with a tweet—and more dildos were thrown afterward.
His point is sharp: the market can create incentives to make the event happen.
Polymarket "created a situation where you could easily make money by betting someone would throw a dildo and then going to a game with a dildo and throwing it."
8.3. Insider trading isn't a bug—it's treated like a feature
John describes suspicious wins that look like inside info:
- A trader made over $1 million betting correctly on Google's most searched people rankings for 2025.
- The same person earned over $150,000 predicting the exact release date of Google's Gemini 3 model.
Then he shows an interview where Polymarket's CEO is asked how they guard against insider trading, and the answer is basically: they don't—because insiders moving the market makes the market "better."
"It creates this financial incentive… for people to go and divulge the information to the market…"
John clarifies: "divulge" meaning people who actually know, and the CEO agrees.
"Yeah."
John bluntly translates what that is:
"You're describing insider trading."
8.4. It gets scarier when the bets are about raids, war, and death
John lists chilling examples where the timing suggests privileged access:
- After the U.S. seized Nicolás Maduro, an anonymous trader made over $400,000, with big bets just hours before the raid was announced.
- In February, two Israelis were charged with using classified military info to place bets.
- In March, anonymous accounts bet that the U.S. would strike Iran by a deadline—many bets placed shortly before bombs fell; one account allegedly wagered $87,000 71 minutes before the strike and won $515,000.
"That sure feels like someone with inside information was betting…"
John notes Kalshi has announced action in two insider trading cases and even brags via billboards that it bans insider trading and doesn't offer "death markets," which John finds suspiciously defensive.
"It'd be like McDonald's putting up ads that its nuggets don't contain baby meat."
Polymarket updated rules saying users can't use stolen confidential information or bet if they can influence outcomes—but John points out suspicious spikes still keep happening, including before a ceasefire announcement that led to hundreds of thousands in winnings.
9. So what can be done—and how should we think about it? 🧭🛑
John says, at minimum, the CFTC should do more, but he doesn't expect much while current leadership remains.
He notes:
- Many cases are moving through courts now.
- It may end up at the Supreme Court.
- Congress has proposals, but meaningful change seems unlikely soon—especially given the president's son advising both major companies.
Still, John argues we need basic guardrails eventually.
In the meantime, he proposes at least changing how we treat these markets culturally—especially how the media platforms them. News organizations should stop "laundering" the companies' reputations by putting odds on screen as if they're news.
"Stop laundering these companies' reputations for them by putting their odds on screen like they are actual news."
He emphasizes a personal-level warning: statistically, most people will lose money, and even if gambling isn't inherently evil, there's something bleak about monetizing every part of life.
"These sites [are] turning every aspect of our lives into a bet."
The deepest loss, he argues, is social: a world where we engage with events as human realities, not just as opportunities to profit.
"A society where things aren't only weighed in financial terms."
He ends with a promise that he won't do anything just because someone bet on it. If he ever starts saying crypto buzzwords, it won't be market manipulation—it'll be a medical emergency.
"If I ever say Bitcoin, Ethereum, blockchain, staking, and web 3… it'll be because I'm having a stroke."
And he closes by circling back to the WNBA story as a comedic escape hatch—joking that if a dildo ever lands on his desk, it won't be because of a secret wager, but because the show needed a funny way out of a long segment.
"A fundamental rule of comedy is: more dildos is more funny."
Conclusion / Final Thoughts
John Oliver's through-line is that prediction markets are being sold as serious financial tools and truth machines, but often function like lightly regulated gambling platforms—sometimes on tragedy and war. He argues the combination of weak enforcement, media partnerships, and incentives for manipulation/insider trading makes them socially corrosive. Until real guardrails arrive, he urges skepticism—especially from news outlets treating odds like facts.
Need a summary like this?
Get instant summaries with Harvest