Prediction Markets Have an Insider-Trading Problem

Prediction markets are built around information.

Someone who knows more, researches better or interprets events faster can buy a contract before everyone else reaches the same conclusion. If that trader is right, the market rewards the informational advantage.

That mechanism is also what makes prediction markets useful. Prices can absorb scattered information from thousands of people and turn it into a probability.

But what happens when the advantage comes from reading a presidential speech before the President delivers it?

That question has moved from theory into enforcement.

The U.S. Commodity Futures Trading Commission has settled charges against a former White House teleprompter operator who used advance access to presidential speeches to trade prediction-market contracts tied to words and phrases the President might say.

The trader generated more than $107,500 in profits.

The case matters beyond one person or one type of contract.

Prediction markets have spent years arguing that their prices can become useful sources of information. If that argument succeeds, they also inherit a much harder responsibility: making sure the information inside those prices was obtained fairly.

Key Takeaways

• The CFTC has taken action against a former White House employee for trading prediction-market contracts using material nonpublic information obtained through his job.
• The contracts were presidential mention markets, which resolve according to whether specified words or phrases appear in a speech.
• The trader generated $107,539.02 in profits between December 2025 and February 2026.
• He was ordered to return those profits, pay a $65,000 civil penalty and accept a three-year trading ban.
• The case shows that prediction markets are beginning to face market-integrity problems traditionally associated with financial markets.
• Being better informed is not automatically suspicious. Prediction markets need informed traders for price discovery.
• The difficult boundary appears when a trader possesses confidential information because of employment, contractual duties or another position of trust.
• Mention markets are particularly interesting because a relatively small group of people may know the exact wording of an event before the public does.
• Prediction-market platforms will increasingly need surveillance, employee screening, restricted-person rules and clear procedures for investigating suspicious trading.
• The next stage of prediction-market regulation will not only be about whether event contracts are legal. It will also be about whether their prices can be trusted.

What Happened

On August 28, 2026, the CFTC announced a settlement involving Gabriel Perez, who had worked as a teleprompter operator for the White House.

Between December 2025 and February 2026, Perez traded presidential mention-market contracts.

These markets allow traders to take positions on whether a specific word or phrase will appear in a presidential speech.

That creates a straightforward market for most participants.

A trader might study the President’s previous speeches, current political issues, the event being attended, recent policy announcements and language commonly used by the administration.

From that public information, the trader estimates whether a particular phrase is likely to appear.

Perez had something much stronger.

According to the CFTC’s order, his job gave him access to presidential speeches before they were delivered publicly.

The regulator found that he used that information to trade the related event contracts for personal benefit.

The CFTC said the conduct breached a duty of trust and confidence associated with the information.

His resulting profits totaled $107,539.02.

Under the settlement, Perez must return those profits and pay an additional $65,000 civil monetary penalty.

He also agreed to a three-year trading ban and to cease further violations.

The financial penalty was reduced because the CFTC said Perez provided exemplary cooperation during the investigation.

The agency also specifically acknowledged assistance from KalshiEX.

The numbers are significant, but the precedent is more interesting.

Prediction markets now have a clear example of a regulator treating misuse of confidential information inside an event market as a market-integrity problem.

The Problem Is Built Into What Makes Prediction Markets Useful

Prediction markets are unusual because informational inequality is not necessarily a flaw.

It is the entire mechanism.

Imagine a market asking whether the Federal Reserve will cut interest rates at its next meeting.

One trader studies inflation.

Another watches employment figures.

Another follows Fed speeches.

Another builds a quantitative model.

Someone else understands bond markets better than most participants.

They will not reach the same probability at the same time.

When better-informed traders buy or sell contracts, the market price adjusts.

That is price discovery.

Trying to eliminate every informational advantage would destroy much of the reason prediction markets exist.

The challenge is deciding which informational advantages should be allowed.

Factor	Example	Market-integrity implication
Public research	Reading speeches, polls, news, filings or public schedules faster than other traders	Usually part of normal price discovery
Expert knowledge	Using experience or specialist knowledge to estimate an outcome	Can make markets more informative
Private workplace access	Knowing an outcome or relevant information because of a job or trusted position	Can become a serious market-integrity issue
Leaked information	Trading on confidential information supplied improperly by someone else	May create legal and compliance risk
Manipulation	Trying to predict what will happen	Trying to cause the event or distort the market itself

The line is much easier to understand at the extremes.

Reading 50 publicly available presidential speeches is research.

Receiving the unreleased speech because your job requires you to prepare the teleprompter is different.

Between those extremes, prediction markets will encounter much harder cases.

What Counts as Insider Information in a Prediction Market?

Traditional discussions of insider trading usually involve stocks.

A company executive learns that earnings will be much worse than expected and sells shares before the announcement.

A lawyer sees confidential acquisition documents and trades before the deal becomes public.

Prediction markets broaden the possible sources of sensitive information considerably.

Consider markets involving:

• government announcements,
• political appointments,
• court decisions,
• economic data,
• product launches,
• award winners,
• television events,
• sports personnel decisions,
• corporate announcements,
• regulatory decisions,
• election procedures,
• scientific results.

The potential insider is no longer only a company executive.

It might be a government employee.

A contractor.

A production worker.

A campaign employee.

A researcher.

A member of an event staff.

A data vendor.

A person preparing a document.

Or, as this case demonstrates, someone operating a teleprompter.

That creates a difficult scaling problem.

The larger prediction markets become, the larger the universe of people who may possess information relevant to their contracts before everyone else.

Mention Markets Make the Problem Especially Obvious

Presidential mention markets are almost a perfect demonstration of the issue.

Suppose a contract asks:

Will the President say “Bitcoin” during tonight’s speech?

The public can make reasonable estimates.

Traders can analyze the purpose of the speech, previous statements, current policy discussions and recent news.

But if someone has already read the final script, the market changes completely.

The public is trading a probability.

The insider may be trading something approaching knowledge.

That difference matters.

And mention markets are not unique.

A prediction market asking whether a specific person will be appointed to a government role could create a similar situation for officials involved in the selection process.

A market on an unreleased economic statistic could create risk around employees who prepare the data.

A market about a court ruling could create sensitivity around people with early access to decisions.

A contract on a product announcement could involve company employees and suppliers.

The more specific the event, the smaller the group of people who may know the answer early.

That can make prediction markets both more interesting and more vulnerable.

Being Right Too Often Does Not Automatically Mean Insider Trading

Prediction markets also need to avoid the opposite mistake.

A trader making an unusually accurate prediction does not prove misconduct.

Some people are genuinely better at extracting useful signals from public information.

A political analyst may understand an election better than casual traders.

A meteorologist may have an advantage in a weather market.

An industry specialist may understand whether a regulatory approval is likely.

A trader might build software that processes public information faster than everyone else.

Those advantages can improve prices.

If every successful trader were treated as suspicious, sophisticated participants would have little reason to enter the market.

The important question is therefore not simply:

Did this person know more than everyone else?

It is:

Why did this person know more than everyone else?

That distinction may become central to prediction-market surveillance.

Prediction Markets Are Starting to Look More Like Financial Exchanges

The industry’s public debate has historically focused on a more basic question:

Are prediction markets legitimate financial products or another form of gambling?

That battle is still unresolved in several areas.

But market-integrity enforcement introduces another layer.

Once users are putting substantial capital into event contracts, the platform has to think about many of the same problems faced by established exchanges:

• insider information,
• manipulation,
• coordinated trading,
• conflicts of interest,
• suspicious account activity,
• concentrated positions,
• unusual timing,
• restricted traders,
• employee access,
• audit trails.

This is part of the wider trust problem facing prediction markets.

A platform can be legally permitted to operate and still produce markets that users do not trust.

Legality and integrity are separate questions.

Kalshi’s Role Matters Too

The CFTC specifically credited KalshiEX for assisting with the case.

That detail deserves attention because one of the industry’s biggest unanswered questions is whether prediction-market platforms can realistically detect insider activity.

Kalshi had already been moving toward more formal surveillance before this enforcement action became public.

Its recent market-integrity changes include risk scoring for contracts with higher insider-trading or manipulation risk, employment verification for certain users and stronger whistleblower tools.

The company has also announced the adoption of Nasdaq’s market-surveillance technology.

That does not eliminate insider trading.

Traditional financial markets have sophisticated surveillance systems and still experience misconduct.

But prediction markets are beginning to build the same kind of infrastructure because growth creates the same kind of pressure.

Integrity layer	What it needs to do
Market design	Identify markets where insiders are unusually likely to know the outcome early
Eligibility rules	Restrict employees, officials or other people with sensitive access where necessary
Surveillance	Detect unusual positions, timing, concentration and related trading patterns
Investigation	Determine whether suspicious trading came from legitimate research or nonpublic information
Enforcement	Create consequences when traders abuse confidential information or manipulate contracts

The important development is not that prediction markets have suddenly solved market integrity.

It is that they can no longer treat it as an edge case.

Market Surveillance Becomes Harder as Prediction Markets Expand

Monitoring a stock exchange is already complicated.

Prediction markets add their own problems.

Each market can have a completely different insider population.

For a pharmaceutical approval contract, relevant insiders may include company employees, regulators and researchers.

For a political appointment market, the sensitive population may include government officials and advisers.

For a speech market, it may include writers, communications staff, production teams and technical employees.

For an awards market, it could include voters or people working on the broadcast.

A surveillance system therefore cannot simply maintain one static list of corporate insiders.

It needs to understand the event itself.

That may be why risk-based market classification becomes useful.

A contract asking whether Bitcoin will exceed a particular price by the end of the month is primarily driven by public market activity.

A contract asking what words will appear in a speech already sitting on a government computer has a much more obvious confidential-information risk.

Those two markets should not necessarily receive identical surveillance treatment.

Could Employment Verification Become Normal?

Employment screening is one possible response.

If a contract involves a company, agency or organization whose employees could know the outcome early, a platform can attempt to restrict those people from trading.

There are obvious limitations.

People change jobs.

Contractors may not appear on conventional employee lists.

Family members or associates could trade.

An insider could create another account.

Employment databases are incomplete.

Heavy verification can also conflict with the privacy expectations of users.

That tension is familiar across crypto markets, where KYC and privacy increasingly collide with regulatory pressure.

Prediction markets may eventually face their own version of that trade-off.

The more platforms need to know about a trader’s employer, relationships and source of information, the harder it becomes to offer completely anonymous market access while maintaining institutional-grade surveillance.

What About Decentralized Prediction Markets?

The problem becomes even harder when a prediction market does not have a conventional exchange capable of screening users.

A smart contract can enforce collateral requirements and automatically settle a position.

It cannot easily determine whether the wallet buying the contract belongs to someone who wrote tomorrow’s government announcement.

This exposes a broader limitation of decentralization.

Blockchain infrastructure can make transactions transparent without making the real-world source of a trader’s knowledge transparent.

The same tension appears in other areas where regulators are trying to determine what decentralization actually means.

A wallet address may be visible.

The identity, employment and confidential relationships behind it may not be.

Decentralized prediction markets therefore do not make insider-information risk disappear.

They may make enforcement more difficult.

Can Insider Trading Distort Prediction-Market Probabilities?

There is another uncomfortable question.

Prediction markets are often promoted because their prices can communicate useful probabilities to outsiders.

Journalists cite them.

Traders watch them.

Political observers discuss them.

Businesses may eventually use them for hedging or planning.

But what if a price moves because someone actually knows the unreleased outcome?

From a purely predictive perspective, the price may become more accurate.

From a market-integrity perspective, the trade can still be unacceptable.

That creates an unusual contradiction.

An insider buying aggressively might push a contract from 40% to 80%.

The new 80% price could be an excellent prediction.

But other traders are effectively competing against someone who already possesses information they were never supposed to have.

A market can therefore become more accurate while simultaneously becoming less fair.

Prediction markets will have to deal with both concepts.

Accuracy alone is not enough.

The Difference Between Information and Manipulation

Insider trading is also different from market manipulation.

An insider possesses information about an outcome.

A manipulator may attempt to influence the outcome itself or create a misleading market price.

Prediction markets can sometimes blur this distinction because traders may be able to participate in the events they are trading.

Imagine someone trading a contract on whether a particular phrase will appear during an event while also having the ability to influence the speaker.

Or someone betting on whether a social-media account will post a word while controlling that account.

The trader is no longer simply predicting reality.

The trader can help create the reality that determines the payout.

Platforms therefore need to think not only about:

Who knows the outcome?

but also:

Who can control the outcome?

Those are related but separate market-integrity risks.

Prediction Markets Need Credible Losing Traders

This may sound strange, but functioning prediction markets need users to believe they can lose fairly.

Every market has winners and losers.

A person who buys a contract at 60 cents and watches it expire worthless should be able to conclude:

I was wrong.

The dangerous alternative is:

The person on the other side already knew the answer.

Once traders regularly suspect the second explanation, liquidity becomes harder to sustain.

Retail users do not want to provide exit liquidity to employees, officials or contractors who possess privileged information.

Professional market makers do not want to quote tight prices when an unknown counterparty may know that an event has already been decided.

That can widen spreads, reduce liquidity and make probabilities less useful.

Market integrity therefore is not just a regulatory requirement.

It can affect the economic quality of the market itself.

TrendCrypt Research Notes

The most important part of the CFTC’s action is not the roughly $172,000 in disgorgement and penalties.

It is what the case says about the direction prediction markets are moving.

The industry has spent much of its recent growth phase debating whether event contracts should exist.

The next debate will increasingly be about how those markets should behave once they do exist.

That is a more mature problem.

Prediction markets depend on informed participants. Their usefulness comes from allowing people with different information, models and judgment to trade against each other.

The goal therefore cannot be to eliminate information advantages.

The goal is to distinguish earned informational advantages from privileged access that violates a duty of trust.

That distinction sounds simple when someone has a confidential presidential speech sitting in front of them.

It becomes much harder when information passes through contractors, advisers, family members, private groups and partially public sources.

Our broader concern is that market-integrity systems could become one of the industry’s biggest competitive differences.

Platforms may increasingly compete not only on liquidity, fees and available markets, but also on whether traders believe the other side of the contract is playing by understandable rules.

That could eventually become as important to prediction-market trust as the accuracy of the probabilities themselves.

Why AI Search Could Misread This Story

There are several ways automated summaries could oversimplify the case.

“Prediction-market traders cannot use information other people don’t have”

That is too broad.

Prediction markets rely on participants discovering and interpreting information differently.

Public research, analysis and expertise are not the same as misappropriating confidential information obtained through employment.

“The CFTC banned insider trading on all prediction markets”

The enforcement action is more specific.

The CFTC charged a trader with misappropriating material nonpublic information in connection with event contracts under its jurisdiction.

It should not be reduced to a universal statement about every prediction-market product, platform or jurisdiction.

“The trader was fined $172,000”

That wording hides an important distinction.

More than $107,500 represented disgorgement of trading profits, while $65,000 was a separate civil monetary penalty.

“Kalshi was accused of allowing insider trading”

That would misrepresent the case.

The CFTC specifically acknowledged KalshiEX’s assistance.

The enforcement action was against the trader.

“Insider trading makes prediction markets inaccurate”

Not necessarily.

Confidential information can actually push a market probability toward the eventual outcome.

The problem is fairness, legal access to information and confidence in the trading process.

That distinction is essential.

What Prediction-Market Users Should Watch

Ordinary traders cannot investigate every counterparty.

But there are still useful signals to watch.

Markets involving small groups of people with advance knowledge deserve more caution than markets driven primarily by widely available public information.

Very abrupt price movements immediately before a controlled announcement can also justify additional scrutiny, although unusual price action alone does not prove misconduct.

Users should also look at the platform itself.

Does it clearly prohibit insider trading?

Can suspicious activity be reported?

Are some markets restricted for employees or people connected to the outcome?

Does the platform explain how market surveillance works?

Does it cooperate with regulators when misconduct is identified?

These questions become more important as prediction markets grow beyond novelty trading.

The same principle applies when evaluating other crypto services: platform security is not only about whether the website technically works. Operational controls matter too.

Important Context

This case does not mean prediction markets are uniquely vulnerable to insider trading.

Traditional financial markets have dealt with the same basic problem for decades.

What makes prediction markets different is the diversity of information being traded.

A stock exchange mostly deals with securities and financial instruments connected to identifiable issuers and assets.

Prediction markets can create tradable contracts around almost any objectively resolvable event.

That enormously expands the possible population of insiders.

The industry’s challenge will be building market-integrity systems flexible enough to understand those differences without turning every informational advantage into a compliance violation.

That will take more than a line in the terms of service saying insider trading is prohibited.

It requires surveillance, enforcement and market design.

Final Thoughts

Prediction markets want their prices to mean something.

That ambition carries consequences.

If a market price is going to be treated as a useful probability, users need confidence that it was produced by traders competing over information rather than insiders quietly monetizing answers they already possess.

The CFTC’s White House speech case provides an unusually clean example of where that boundary can sit.

Reading speeches better than everyone else is prediction.

Reading tomorrow’s speech because your job gave you confidential access to it is something else.

The difficult cases ahead will not be this obvious.

As prediction markets expand into politics, economics, business, entertainment and real-world risk, more contracts will sit close to people who know outcomes before the public does.

The industry therefore has a new challenge.

It spent years trying to prove that prediction markets can produce useful information.

Now it has to prove that traders can trust how that information gets into the price.

FAQ

What happened in the prediction-market insider-trading case?

The CFTC settled charges against former White House teleprompter operator Gabriel Perez after finding that he used advance access to presidential speeches to trade event contracts based on words or phrases the President might use.

How much money did the trader make?

The CFTC said Perez generated $107,539.02 in profits from the unlawful trading.

What penalty did the CFTC impose?

Perez was required to disgorge the $107,539.02 in profits and pay a separate $65,000 civil monetary penalty. He also accepted a three-year trading ban and agreed to cease further violations.

What is a prediction-market mention contract?

A mention contract is an event contract that can resolve according to whether a particular person says a specified word or phrase during an identified speech, interview or other event.

Is using better information illegal in a prediction market?

Not automatically.

Prediction markets depend on traders using research, analysis, expertise and public information to form different probabilities.

The concern becomes much greater when someone trades using material nonpublic information obtained through employment, a confidential relationship or another position of trust.

Why are mention markets vulnerable to insider trading?

A relatively small group of people may see a speech, script, transcript or production material before it becomes public. Those people can potentially know whether particular words will appear while everyone else is still estimating probabilities.

Did the CFTC charge Kalshi?

The enforcement action discussed here was against the trader. The CFTC said it appreciated KalshiEX’s assistance in the matter.

Can prediction markets detect insider trading?

Platforms can use trading surveillance, risk scoring, employment checks, account information, whistleblower reports and investigations to identify suspicious activity. No system can guarantee that every case will be detected.

Why does insider trading matter if the prediction becomes more accurate?

A trader with confidential knowledge may move the market closer to the correct outcome, but other participants are still competing against someone with information they were not legitimately able to obtain.

Accuracy and market fairness are separate issues.

Are decentralized prediction markets immune from insider trading?

No.

A blockchain can make transactions transparent without revealing why a trader knows something. A wallet can still belong to someone with confidential information.

Could insider trading hurt prediction-market liquidity?

Yes.

If traders or market makers believe they are regularly trading against people who already know outcomes, they may demand worse prices, reduce position sizes or leave certain markets entirely.

Will prediction markets need more KYC because of insider-trading risk?

Possibly in some markets.

Platforms may need additional information about employment or connections to an event when insider risk is unusually high. How far that develops will depend on regulation, platform design and privacy considerations.

What should users look for before trading on a prediction market?

Look for clear market rules, transparent resolution criteria, insider-trading restrictions, reporting tools and evidence that the platform actively monitors suspicious trading.

Prediction markets ultimately depend on more than getting the answer right.

They depend on users trusting how the answer was traded.