ZachXBT Named Axiom in a $40 Million Insider Trading Market — Then 12 Wallets Insider-Traded the Market Itself

On February 26, 2026, blockchain investigator ZachXBT published what may become the most consequential expose in memecoin trading history: audio recordings, chat logs, dashboard screenshots, and on-chain evidence proving that employees of Axiom Exchange — a Solana-based trading platform backed by Y Combinator — used internal surveillance tools to front-run their own users’ wallets for profit. Within hours of the teaser announcement, on-chain analysts proved that a $39.7 million prediction market on Polymarket, created to guess which company ZachXBT would name, was itself insider-traded by at least 12 wallets that collectively turned roughly $400,000 into $1.02 million. The largest single trader, operating under the handle “predictorxyz,” converted $65,800 into $411,400 — an approximately 7x return — using a wallet that was freshly created and never touched any other market. The same week, Kalshi announced its first-ever enforcement actions against insider traders on its platform, and the Commodity Futures Trading Commission issued its inaugural prediction market advisory, signaling that regulators are no longer content to observe from the sidelines as insider trading becomes a recurring feature of decentralized prediction markets.

The Axiom Investigation: Audio Recordings, Wallet Maps, and a $200,000 Plan

Axiom Exchange is a Solana-based memecoin trading platform that emerged from the Y Combinator Winter 2025 batch, co-founded by Henry Zhang and Preston Ellis. In a market defined by fleeting hype cycles and razor-thin margins, Axiom distinguished itself by building infrastructure that made high-frequency memecoin trading accessible to retail users — and enormously profitable for the platform itself. The company has generated over $390 million in revenue, a figure that positions it as one of the most commercially successful tools in the memecoin trading space. Its rapid rise was fueled by the explosive growth of Solana-based memecoins throughout 2025 and into 2026, a period during which billions of dollars flowed through decentralized exchanges and trading bots. But beneath the surface of that commercial success, a small group of employees were allegedly exploiting the very users who made it possible.

At the center of ZachXBT’s investigation are the internal dashboards that Axiom provided to certain employees — tools that exposed the full wallet lists of every Axiom user, complete transaction histories, wallet nicknames, linked accounts, and the ability to track any user’s trading activity in real time. These dashboards were ostensibly designed for operational purposes: customer support, fraud detection, and platform analytics. But according to ZachXBT’s evidence, they were weaponized for personal profit by employees who recognized that knowing what high-profile traders were about to buy gave them a license to print money. In the memecoin market, where tokens can move 50% or more on a single influential trade, front-running — buying a token milliseconds before a large order pushes the price up, then selling into the resulting spike — is extraordinarily lucrative. The dashboards gave Axiom employees something no external trader could replicate: a real-time window into the intentions of every user on the platform.

The most damning piece of evidence is an audio recording published by ZachXBT in which Broox Bauer, identified as a senior business development employee at Axiom, explicitly describes the scope of his access and his plans to monetize it. In the recording, Bauer discusses how he could track any Axiom user and obtain detailed information about their trading behavior, wallet connections, and personal details. He outlines what he calls “ground rules” for a coordinated trading operation and describes a $200,000 plan involving Gowno, also known as Seb, another individual named in the investigation. The audio recordings are not ambiguous or open to charitable interpretation — they capture a company insider describing, in his own words, a systematic scheme to exploit privileged access to user data for financial gain. ZachXBT supplemented the audio with chat logs and dashboard screenshots that corroborate Bauer’s claims and demonstrate the operational reality of the front-running scheme.

“I can track any Axiom user… find out anything to do with that person.”

Broox Bauer, in audio recording published by ZachXBT

Beyond Bauer, ZachXBT named three additional individuals connected to the operation. Ryan, known online as Ryucio, and Gowno, known as Seb, were identified as active participants in the front-running scheme. A fourth individual, referred to only as “Mystery,” was also named but with fewer details provided about their specific role. The investigation further uncovered a Google Sheet listing crypto Key Opinion Leaders — commonly known as KOLs — who were specifically targeted for front-running. These are high-profile traders and influencers whose publicly visible or internally observable trades move markets. One KOL identified on the list was referred to as “Marcell.” The operation was not opportunistic or ad hoc; it was organized, documented, and designed to systematically extract value from the traders whose activity generated the largest market impact on the Axiom platform.

ZachXBT’s on-chain tracing of Bauer’s primary wallet revealed fund flows to five separate centralized exchange deposit addresses, a pattern consistent with systematic extraction and off-ramping of profits. The use of multiple CEX deposit addresses suggests an attempt to distribute withdrawals across different accounts or exchanges, potentially to reduce the visibility of any single large inflow. ZachXBT noted that Bauer is based in New York City, placing him within the jurisdiction of the Southern District of New York — the same federal district that prosecuted the Mango Markets manipulation case, the Eisenberg case, and other landmark crypto fraud cases. The SDNY has established itself as the most aggressive federal jurisdiction for crypto enforcement, and Bauer’s geographic proximity to that office carries significant legal implications. Axiom’s official response was swift but measured, characterizing the company’s leadership as blindsided by the conduct of what it described as a former employee.

“We are shocked and disappointed by the actions of a former employee. Dashboard access has been revoked and we are cooperating with ongoing investigations.”

Axiom Exchange, official statement

The $39.7 Million Prediction Market That Became the Story

The saga began not with the investigation itself but with a teaser. On February 23, 2026, ZachXBT posted a cryptic message on X announcing that he was preparing to publish a major investigation into a crypto company — without naming which one. The post was deliberately vague, designed to build anticipation while protecting the integrity of the forthcoming publication. It worked beyond anything he likely anticipated. The teaser garnered over 11 million views and 27,000 likes, transforming a routine pre-announcement into a crypto-wide guessing game. Every major crypto community, from Discord servers to Telegram groups to X threads, erupted with speculation about which company would be named. Within hours, that speculation migrated to the one place where opinions could be monetized: Polymarket.

A prediction market was created on Polymarket offering 28 possible outcomes, each representing a different crypto company that could be the target of ZachXBT’s investigation. The market attracted $39.7 million in total volume, making it one of the largest single-event prediction markets ever hosted on the platform and underscoring just how much attention and capital ZachXBT’s work commands in the crypto ecosystem. Traders could buy shares in any outcome they believed was correct, with prices fluctuating in real time based on the aggregate weight of money flowing into each option. The market functioned as a real-time barometer of collective belief — or, as it turned out, a barometer of who had insider knowledge and who was guessing.

For most of the market’s active life, Meteora emerged as the overwhelming crowd favorite. The Solana-based decentralized exchange had been embroiled in a separate controversy over the TRUMP and MELANIA memecoins launched in early 2025, with allegations that insiders had extracted approximately $1.2 billion from retail traders through coordinated pump-and-dump operations. A lawsuit filed by Benjamin Hurlock added legal weight to the accusations, and the general sentiment in crypto circles was that Meteora was the most likely target for an investigator of ZachXBT’s caliber. Meteora’s probability peaked between 47% and 53%, commanding more volume than any other single outcome. The reasoning was logical: Meteora had the biggest scandal, the most public controversy, and the clearest trail of alleged wrongdoing. But the crowd was wrong.

In the final hours before ZachXBT’s publication, the Axiom outcome surged from low single-digit percentages to 46.2% — a dramatic, unmistakable shift that, in retrospect, was driven almost entirely by insiders who knew the answer before anyone else. The timing of these bets was not subtle. Large positions were opened in rapid succession, each one pushing the Axiom probability higher and attracting attention from other traders who interpreted the price movement as a signal. Some of those followers placed copycat bets, further inflating the Axiom outcome and amplifying the insiders’ returns. The market, designed to aggregate distributed information into accurate probabilities, instead became a vehicle for those with privileged knowledge to profit at the expense of everyone else.

The episode illustrates the fundamental tension at the heart of prediction markets as an information technology. When they work as intended, they are among the most powerful tools for aggregating dispersed knowledge into actionable probabilities — a function that has attracted support from economists, technologists, and policymakers alike. But when participants possess information obtained through insider access rather than superior analysis or legitimate research, the mechanism becomes an extraction tool rather than an information aggregator. The ZachXBT-Axiom market is a case study in both possibilities, and it has reignited the debate over whether prediction markets can function fairly without enforceable rules against insider trading. For a broader look at how this fits into the competitive landscape, see the prediction market industry overview.

Twelve Wallets, $1 Million in Profit

Within hours of ZachXBT’s publication, on-chain analysts descended on the Polymarket contract data to determine who had profited from advance knowledge of the investigation’s target. Lookonchain, one of the most widely followed blockchain analytics accounts, was first to publish a comprehensive breakdown. Their analysis identified 12 wallets that collectively invested approximately $400,000 in the Axiom outcome and extracted $1.02 million in profit — a staggering return that far exceeded what any legitimate analysis could have produced in the compressed timeframe between ZachXBT’s teaser and his final publication. The pattern across these wallets was remarkably consistent: large, concentrated bets placed on the Axiom outcome at a time when the broader market still overwhelmingly favored Meteora, followed by complete withdrawal of funds immediately after resolution.

DefiOasis, another prominent on-chain researcher, provided additional granularity. Their analysis revealed that 8 of the top 10 earners on the entire ZachXBT prediction market were insiders who possessed advance knowledge of the investigation’s target. This means that the leaderboard for the market — the list of traders who profited the most — was almost entirely composed of individuals who were not trading on analysis, intuition, or crowd wisdom, but on private information. WuBlockchain contributed further data, reporting that over 3,630 unique addresses traded on the market in total. On the losing side of the ledger, 52 addresses lost between $10,000 and $100,000 or more, representing retail traders who were betting against insiders without knowing it. The asymmetry is stark: a small number of informed wallets extracted outsized profits while thousands of uninformed participants absorbed the losses.

The most scrutinized wallet belongs to the user identified as “predictorxyz.” This wallet invested $65,800 in the Axiom outcome and withdrew $411,400 — an approximately 7x return that represents the single largest documented profit from the market. What makes the predictorxyz wallet particularly damning is its behavioral profile: the wallet was newly created, had no prior trading history on Polymarket, and only ever interacted with the Axiom outcome. It never touched any of the other 27 markets available. This is not the behavior of a sophisticated trader diversifying across outcomes or hedging positions — it is the behavior of someone who knew the answer and created a wallet specifically to capitalize on that knowledge. The wallet’s creation timing, betting pattern, and immediate withdrawal after resolution form an almost textbook case of insider trading conducted through pseudonymous infrastructure.

ZachXBT’s own investigation of the Polymarket insider activity added another layer. He traced one wallet to a “JustADegen” Solana wallet through instant exchange services — the kind of cross-chain swapping tools that allow users to move funds between blockchains without centralized intermediaries. The use of instant exchange services suggests a deliberate attempt to obscure the connection between the trader’s Polymarket activity on Polygon and their primary identity on Solana, where they may be known to the crypto community. Two other wallets were created just three hours before ZachXBT’s publication, investing a combined $59,800 and withdrawing approximately $109,000. The three-hour window is significant because it falls after ZachXBT would have finalized his investigation and likely shared it with a small number of trusted contacts for review — creating a narrow window in which someone with advance access could act.

The Agash wallet represents another documented case: a $50,000 bet that timing analysis connected to an individual with links to the investigation. Separately, an anonymous trader placed a $50,700 bet on Axiom that, due to its size and timing, triggered a wave of copycat bets from other traders monitoring the market for large-order flow. This copycat effect is a well-documented phenomenon in prediction markets, where uninformed traders interpret large bets as signals and pile in accordingly — effectively allowing the insider to amplify their own trade’s impact on the market price. Polysights, another analytics firm, identified a cluster of 5 wallets that collectively turned approximately $50,000 into $266,000, exhibiting coordinated timing and behavioral patterns consistent with a single operator or small group.

CoinDesk’s reporting on the incident captured what may be the most important analytical observation about the entire episode. The structural irony, as CoinDesk framed it, is that the prediction market mechanism worked exactly as its proponents claim it should — it aggregated private information into prices with remarkable speed and accuracy. The Axiom outcome surged in the final hours precisely because people with real information were putting real money behind it. The problem, of course, is that the information was obtained not through superior analysis, independent research, or legitimate reporting, but through insider access to a forthcoming investigation. The market did not fail as an information aggregation tool; it succeeded, but in a way that enriched insiders at the expense of the uninformed majority. This distinction is at the heart of the regulatory and philosophical debate now engulfing prediction markets.

“The structural irony here is that the mechanism worked exactly as designed. Prediction markets are supposed to aggregate private information into prices. The problem is when that private information was obtained through insider access rather than superior analysis.”

CoinDesk

The Kalshi Counterpoint: $20,397 Fines and a 2-Year Ban

In a timing coincidence that could not have been more pointed, Kalshi — Polymarket’s primary regulated competitor — announced its first two public enforcement actions on February 25, 2026, exactly one day before ZachXBT published the Axiom investigation. The juxtaposition was devastating for Polymarket’s credibility. While Polymarket was about to be engulfed in yet another insider trading scandal with zero enforcement infrastructure to address it, Kalshi was demonstrating that a prediction market platform could identify, investigate, and punish insider trading without waiting for regulators to act first. The two enforcement actions were modest in scale but enormous in symbolic weight, establishing a precedent that prediction market platforms can and should police their own markets.

The first enforcement action targeted Artem Kaptur, who was fined $20,397 and banned from the platform for two years. Kaptur had traded on markets related to MrBeast streaming events while possessing non-public information about upcoming streams — information that gave him a material advantage over other participants who were making probabilistic assessments based on publicly available data. The second action targeted Kyle Langford, who was fined $2,246 and banned for five years — a longer ban despite a smaller fine, reflecting the severity of his conduct. Langford had bet $200 on himself in a market related to his own activities, a textbook case of insider trading so straightforward that it would be comic if it were not a genuine breach of market integrity. The five-year ban signals that Kalshi views self-dealing as a particularly egregious violation, regardless of the dollar amount involved.

Behind these enforcement actions lies an infrastructure that Kalshi has been building quietly for months. The platform operates an internal surveillance system called Poirot — named after Agatha Christie’s fictional detective — which monitors trading patterns for anomalies consistent with insider trading, market manipulation, and other forms of misconduct. Kalshi has disclosed that it has conducted over 200 investigations using this system, meaning the two public enforcement actions represent the tip of a much larger compliance iceberg. The platform maintains a dedicated audit committee, and it recently hired Robert DeNault to lead its enforcement division, a move that signals institutional seriousness about market integrity. CEO Tarek Mansour drew explicit and unmistakable contrasts with Polymarket’s approach, positioning Kalshi as the platform that takes market integrity seriously in a space where its largest competitor appears to treat insider trading as a feature rather than a bug.

Luana Lopes Lara, Kalshi’s co-founder, was characteristically blunt in her public commentary on the enforcement actions, stating simply: “F***ed around, found out.” The comment, while informal, captured the essence of Kalshi’s message: that prediction market participants should expect consequences for trading on non-public information, and that the absence of enforcement on other platforms does not mean such conduct is acceptable. The contrast with Polymarket could not be starker. Kalshi operates with full KYC/AML compliance, explicit prohibitions on insider trading, and a demonstrated willingness to investigate and punish violations. Polymarket operates with pseudonymous wallet-based trading, no disclosed surveillance tools, no enforcement actions in its history, and a CEO who has publicly described insider trading as “cool.” For a detailed analysis of Kalshi’s enforcement framework, see our coverage of Kalshi’s insider trading enforcement.

Polymarket’s Pattern: From Google to Maduro to the Super Bowl

The ZachXBT-Axiom insider trading incident is not an isolated event — it is the latest entry in a growing pattern of insider trading on Polymarket that stretches back to mid-2025 and has accelerated dramatically in recent months. The earliest documented case involves the Israel-Iran conflict markets in June 2025, where a user operating under the handle ricosuave666 earned $152,000 by placing bets that appeared to reflect advance knowledge of military operations before they were publicly reported. That case became historically significant when it led to the first criminal prosecution related to prediction market insider trading, establishing a legal precedent that the abstract concept of “insider trading on prediction markets” could be charged as a real crime with real consequences. For a comprehensive analysis, see our coverage of the Israel-Iran insider trading case.

In December 2025, the scale escalated dramatically with the Google Year in Search incident. A trader operating under the name AlphaRaccoon earned $1.15 million by betting on the correct Google Year in Search results before they were publicly announced. The trader’s positions were placed with surgical precision, targeting outcomes that had no publicly available basis for prediction — the kind of information that could only have come from inside Google or from someone with access to the results before their official release. The case demonstrated that insider trading on prediction markets was not limited to geopolitical events or crypto-native contexts; it could extend to any domain where non-public information existed and a corresponding market was available.

January 2026 brought the Maduro capture market, where a single trader turned $34,000 into $436,000 — a 12x return — by betting on the capture of Venezuelan president Nicolas Maduro before the operation was publicly known. The sheer magnitude of the return, combined with the geopolitical sensitivity of the underlying event, drew attention from mainstream media and intelligence community observers who noted that prediction markets were now functioning as real-time indicators of covert operations. In February 2026, the Super Bowl LX market generated an estimated $2 to $3 million in insider profits from bets placed with apparent advance knowledge of game outcomes or key events, though the specific mechanism of information leakage remains under investigation. For details on the Maduro case, see our analysis of the Maduro capture bets.

What ties all these incidents together is not just their occurrence on the same platform, but Polymarket’s consistent refusal to acknowledge, investigate, or address them. CEO Shayne Coplan has been openly and enthusiastically pro-insider-trading in his public statements. In a 60 Minutes interview and in quotes published by Axios, Coplan stated that what makes Polymarket interesting is the financial incentive it creates for people with information to put their money where their mouths are. This is not a nuanced position dressed in the language of market efficiency — it is an explicit endorsement of the principle that insider trading is a feature of Polymarket’s design, not a bug. Coplan’s position is philosophically coherent within a narrow libertarian framework that views all information asymmetry as legitimate, but it is wildly out of step with every other financial market in existence and with the regulatory expectations that are now converging on the prediction market industry.

Perhaps most telling is what is absent from Polymarket’s Terms of Service. Unlike every other major financial platform — from stock exchanges to commodities markets to even other prediction market platforms like Kalshi — Polymarket’s Terms of Service contain no prohibition on insider trading whatsoever. There is no clause defining it, no language prohibiting it, and no mechanism for enforcement even if the platform wanted to act. This is not an oversight; it is a design choice that reflects Coplan’s publicly stated philosophy. The result is a platform that has now been the site of at least five major insider trading incidents in eight months, totaling millions of dollars in documented insider profits, without a single enforcement action, public investigation, or rule change in response. The pattern is not breaking; it is accelerating.

“What’s cool about Polymarket is that it creates this financial incentive for people who have information to put their money where their mouth is.”

Shayne Coplan, Polymarket CEO

Regulators Close In

The regulatory response to the mounting pattern of insider trading on prediction markets reached a turning point in February 2026, when the Commodity Futures Trading Commission issued its first-ever Prediction Markets Advisory under Release 9185-26. The advisory marks a decisive shift from the observation posture that the CFTC had maintained throughout 2024 and most of 2025, during which the agency appeared content to let the prediction market industry develop before intervening. That period of forbearance is now over. The advisory explicitly addresses insider trading, market manipulation, and fraud on prediction market platforms, and it signals that the CFTC considers these platforms to fall within its regulatory jurisdiction — a position that Polymarket and others in the industry had hoped to avoid. CFTC Chairman Selig issued a direct warning to market participants that left no room for ambiguity about the agency’s enforcement intentions.

“If you attempt to engage in manipulation, fraud, or insider trading on prediction markets, we will find you and we will hold you accountable.”

CFTC Chairman Selig

The CFTC advisory did not arrive in isolation. Jay Clayton, the U.S. Attorney for the Southern District of New York — the same office whose jurisdiction covers Axiom employee Broox Bauer — issued his own public statement that carried even more immediate weight for the individuals named in ZachXBT’s investigation. Clayton’s statement was targeted and unambiguous: a prediction market does not insulate anyone from fraud, and trading on material non-public information will be prosecuted regardless of the platform on which it occurs. The SDNY’s involvement is particularly significant because it is the same office that secured convictions in the Eisenberg Mango Markets case and has built the deepest institutional expertise in crypto-related financial crimes of any federal prosecutorial office in the United States. For individuals like Bauer, who is based in New York City, Clayton’s statement is not abstract policy — it is a direct signal.

“A prediction market doesn’t insulate you from fraud. If you trade on material non-public information, you will be prosecuted.”

Jay Clayton, U.S. Attorney, SDNY

On the legislative front, Representative Ritchie Torres introduced H.R. 7004, the Public Integrity in Financial Prediction Markets Act, which would create a federal framework specifically targeting insider trading on prediction markets. The bill addresses a gap in existing law: while securities insider trading statutes are well-established, their applicability to prediction market contracts — which are not technically securities in most cases — remains legally uncertain. H.R. 7004 would resolve that uncertainty by creating prediction-market-specific prohibitions, enforcement mechanisms, and penalties. The bill has bipartisan support and is advancing through committee at a pace that reflects the political urgency created by the accumulating insider trading scandals. For a detailed analysis of the legislation, see our coverage of the Torres bill.

The international regulatory picture provides further context for the direction of travel. The Netherlands became the first country to ban Polymarket entirely, classifying the platform as illegal gambling under Dutch law and ordering its removal from the market. While the Dutch classification differs from the U.S. approach — which treats prediction markets as financial instruments rather than gambling — the outcome is the same: a major jurisdiction has concluded that Polymarket’s operating model is incompatible with consumer protection standards. The Netherlands ban, combined with the CFTC advisory, the Torres bill, and the SDNY signals, creates a regulatory convergence that is unprecedented in the short history of prediction markets. The industry is no longer in a gray zone; the lines are being drawn, and the message from regulators on both sides of the Atlantic is that the era of unregulated prediction market trading is ending. For background on the Dutch decision, see our analysis of the Netherlands ban.