The logs do not lie, but they also do not tell the full story. Over three months, a single wallet—linked to a White House official—traded on non-public information, and the platform’s response remains a black box. The incident, first reported by a mainstream outlet, centers on Gabriel Perez, a White House employee who repeatedly bought Kalshi contracts tied to President Trump’s public mentions of specific terms—minutes before those mentions occurred. The platform, Kalshi, is a registered CFTC exchange, bound by strict rules against insider trading. Yet the timeline of their detection and restriction is conspicuously absent. This is not a story about a rogue trader; it is a story about the gaps in regulatory oversight that allow ghosts to trade in the silence between blocks.
To understand the gravity, we must first trace the architecture of the case. Kalshi is a novel entity—a prediction market where users trade event contracts, including “mention markets” that pay out based on whether a public figure (like the U.S. President) says a specific word within a given window. These contracts are classified as derivatives under CFTC jurisdiction, making them subject to the same anti-fraud and anti-manipulation rules as commodity futures. In particular, CFTC Rule 6(c)(1) prohibits the use of manipulative or deceptive devices in connection with any swap, contract of sale, or commodity transaction. The agency’s 2022 Advisory explicitly states that insider trading applies to prediction markets, and that exchanges have an independent duty to prevent and report such activity. This is the legal bedrock upon which the scandal rests.
Gabriel Perez, a mid-level White House staffer with access to real-time schedules and speech drafts, allegedly used his position to front-run Kalshi contracts. From March 2025 through early June 2025, he placed dozens of trades—some as large as $5,000—on mentions of words like “inflation,” “infrastructure,” and “Ukraine.” The probability of these events jumped immediately after his orders were filled, and within hours, Trump’s public appearances confirmed the pattern. Kalshi’s surveillance team flagged the activity as anomalous, restricted Perez’s account, and reported him to the CFTC. But here lies the fracture: the platform has refused to disclose the precise timestamps of each action—when the trade was placed, when it was flagged, when the restriction was applied, and when the report was filed. Without these data points, we cannot verify whether further trading occurred between detection and restriction. The numbers hold the memory we ignore.
Mapping the invisible currents of liquidity—this is where on-chain forensics meets regulatory analysis. Although Kalshi is not a blockchain-based platform, its internal logs function as a central ledger, and the absence of temporal granularity is akin to a smart contract missing a timelock. In my years auditing DeFi protocols, I have learned that the most critical element in proving or disproving insider trading is the exact sequence of events. A delay of even 10 minutes can allow a sophisticated actor to alter his position, secure profits, or destroy evidence. In the traditional equities world, the SEC mandates that brokers and exchanges timestamp every order to the millisecond and retain these logs for at least five years. Kalshi, as a registered DCM, must adhere to similar standards under CFTC recordkeeping rules (Regulation 1.31). Their silence on the timestamps suggests either a failure to capture them, or a legal strategy to avoid self-incrimination. Neither option is reassuring.
Let us examine the evidence chain. According to the report, Kalshi’s compliance team “swiftly” identified the pattern after Perez’s first suspicious trade. However, the word “swiftly” is subjective. Did the detection take hours or days? Did the team have to manually review each trade, or was an automated flag triggered? And critically, was Perez able to continue trading after the flag while the review was ongoing? The platform’s new “employment screening” measures, announced on June 9, 2025—three months after the first trade—are reactive, not preventive. They now bar employees of certain government agencies from trading in related contracts. But the list remains undisclosed. Does it cover the White House? The Treasury? The Fed? Silence speaks louder than floor prices.
Root cause forensics reveals a deeper systemic vulnerability. Kalshi’s surveillance model relies on traditional compliance watchlists and pattern recognition, but it lacks the real-time, on-chain auditability that decentralized prediction markets like Polymarket offer. In a smart contract-based market, every trade is permanently recorded with a block timestamp, making it impossible to hide the sequence of events. Here, centralization provides efficiency, but it also creates a single point of opacity. This case exposes that opacity as a liability. If Kalshi cannot prove that it acted within minutes of detecting suspicious activity, then its entire compliance posture is suspect. The CFTC’s upcoming decision on whether to bring an enforcement action against Kalshi—not just Perez—will hinge entirely on those missing timestamps.
The contrarian angle: Is the focus on timestamps a red herring? Some might argue that even if Kalshi restricted Perez within hours, the damage was done—the market moved, and Perez’s profits were locked. The real issue is the mere existence of information asymmetry in a market that claims to be democratized. But I counter: without precise timestamps, we cannot determine if Kalshi’s systems were negligent or merely slow. If the restriction came within 30 minutes of the first trade, perhaps the system worked reasonably well. If it took three days, that is a systemic failure. The difference is critical for regulatory precedent. Moreover, we must consider that Perez’s trades were relatively small—total profits estimated under $100,000. The pattern was clear, but the financial impact was minor. Is the CFTC willing to devote resources to set a precedent on a small case? History suggests they may settle, leaving the timestamp gap unresolved. But for the industry, the longer the uncertainty persists, the more the narrative of “regulated markets are safe” cracks.
Let us turn to the opportunity cost. The same week the scandal broke, Trump Media & Technology Group announced Truth API, a paid service offering millisecond-level access to posts from Truth Social accounts. This is a direct response to the information arbitrage gap. If you can legally buy instant access to public statements, you no longer need insider sources. This creates a new layer of market infrastructure: data-as-a-service for prediction market traders. The irony is poetic—a scandal about illegal non-public information is accelerating the commodification of legal public information. In a bear market, survival matters more than gains, and protocols that can prove their data feeds are both fast and clean will attract liquidity. Truth API is not a blockchain solution, but it demonstrates that the demand for time-sensitive data is skyrocketing.
Watching the block confirm, not the narrative—the next signal for analysts is the CFTC’s response. If the agency releases a public order with detailed timestamps, we will have the first real-world benchmark for acceptable detection and restriction latency. If they remain silent, the precedent is that exchanges can hide behind vague language. I suspect the CFTC will demand more granular reporting from all registered DCMs, similar to the SEC’s Market Information Data Analytics System (MIDAS). This would be a boon for compliance technology startups and a burden for existing platforms. The truth is not in the tweet, but in the transaction.
Coloring the grey areas of market sentiment—where does this leave the average user? The case does not invalidate prediction markets as a tool for price discovery. It simply reminds us that any centralized intermediary introduces counterparty risk. For those willing to accept the tradeoff, Kalshi remains one of the few legal outlets in the U.S. But for the data detective, the takeaway is clear: always audit the audit trail. Ask: when was the flag raised? When was the account frozen? If the platform cannot answer, the ghost is still in the machine.
