On May 27, the U.S. Department of Justice unsealed an indictment against a Google security engineer who allegedly used confidential internal search data to place a series of extraordinarily prescient bets on the prediction market Polymarket, netting approximately $1.2 million in illicit profits. The case marks the second criminal prosecution for insider trading on a decentralized prediction platform and raises urgent questions about how these markets police themselves when participants possess material non-public information.
The Man Behind the AlphaRaccoon Account
Federal prosecutors charged Michele Spagnuolo, a 36-year-old Italian national residing in Switzerland, with three criminal counts: commodities fraud, wire fraud, and money laundering. While the indictment identifies him as a software engineer at Google, his professional profile reveals a far more specialized role. According to his LinkedIn history, Spagnuolo spent more than 12 years in Google’s information security division, where he helped build the company’s web security infrastructure and co-authored the CSP3 specification — a foundational standard for content security on the web.
The government alleges that in May 2024, Spagnuolo opened a Polymarket account under the handle “AlphaRaccoon.” Between October 15 and December 4, 2025, he placed at least 23 wagers totaling approximately $2.75 million on prediction markets tied to Google’s annual “Year in Search” ranking — the company’s year-end list of the most-searched terms and people globally.
How the Scheme Unfolded
According to the indictment, Spagnuolo accessed non-public search trend data through an internal Google tool that only employees can reach. The tool’s interface displayed a “Google Confidential” watermark in red — a detail prosecutors emphasized to establish that the defendant knew the information was proprietary. Spagnuolo had also signed confidentiality agreements with the company.
What did Spagnuolo bet on? The markets asked questions such as “Who will be the most searched person on Google this year?” Armed with the internal data, he knew the answer before the public did — and before most other Polymarket participants had any reliable basis for their own wagers.
The d4vd Bet That Made Everything Obvious
The most telling wager involved the singer d4vd. At the time Spagnuolo placed his bet, Polymarket traders rated d4vd’s chance of becoming the most-searched person of 2025 as virtually zero. The market assigned a negligible probability to that outcome because no public data suggested d4vd would top the list.
On December 4, 2025, Google published its Year in Search 2025 results. d4vd had indeed ranked first. The AlphaRaccoon account instantly gained approximately $1.2 million in profit. Independent analysts later calculated that 22 of the 23 bets placed by the account had landed correctly — a statistical outcome so improbable that it drew immediate suspicion across the Polymarket community.
“Spagnuolo accessed Google’s confidential and commercially valuable internal data and knew the outcome of the bets before other market participants,” the indictment states. “Those on the other side of his trades did not have access to that information.”
Suspicion Emerged the Same Week
The unnatural winning streak did not go unnoticed. Within days of the Year in Search release, multiple Polymarket users publicly flagged the AlphaRaccoon account as a likely insider trader. Online forums lit up with analyses of the wallet addresses, the timing of the bets, and the improbably high accuracy rate.
As scrutiny intensified, prosecutors allege, Spagnuolo attempted to cover his tracks. He changed his Polymarket handle, moved funds through cryptocurrency privacy services, and took steps to obscure the origin and ownership of the money. These actions form the basis of the money laundering charge.
Google’s Response and the Confidentiality Question
Google placed Spagnuolo on leave after learning of the investigation and issued a statement emphasizing that while the internal search trends tool is accessible to employees as part of their work, using that access to place personal bets constitutes a severe policy violation. The company noted that the tool’s interface carried a visible “Google Confidential” classification — making it clear that the data was never intended for external use, let alone for gambling on prediction markets.
What Are the Charges and What Does Spagnuolo Face?
The indictment charges Spagnuolo with three federal crimes. Commodities fraud stems from the allegation that he used material non-public information to trade on a derivatives platform — Polymarket’s contracts qualify as commodity interests under the Commodity Exchange Act. Wire fraud covers the use of electronic communications to execute the scheme. Money laundering addresses his efforts to conceal the proceeds.
If convicted on all counts, Spagnuolo faces a maximum statutory sentence of 50 years in prison. He was arrested in New York on May 27, appeared before a federal magistrate, and was released on a $2.25 million bond. He has not yet entered a plea.
This Is the Second Polymarket Insider Trading Case in Two Months
Spagnuolo’s prosecution follows closely on the heels of a similar case. In April 2026, the DOJ charged a U.S. special operations soldier with using classified information about the Venezuelan government’s capture of former president Nicolás Maduro to place bets on Polymarket, earning more than $400,000. That defendant has pleaded not guilty.
The two cases share a common thread: both defendants allegedly had privileged access to information about events that directly translated into prediction market outcomes. In the soldier’s case, the information came from operational planning. In Spagnuolo’s case, it came from a Google internal dashboard.
Polymarket’s Structural Vulnerability
Polymarket has publicly stated that its own fraud detection systems flagged the AlphaRaccoon activity and that it cooperated with federal investigators and the CFTC. The company revised its terms in March 2026 to explicitly prohibit bets based on stolen confidential information and bets placed by individuals who can influence the outcome of the events they wager on.
Yet the two insider trading cases expose a deeper structural issue. Prediction markets derive their value from the “wisdom of the crowd” — the idea that aggregated public knowledge produces accurate forecasts. But when participants with inside information enter the pool, the crowd is no longer wise. It is playing against someone who already knows the answer.
Stock markets have spent decades building legal and regulatory frameworks to handle this problem. Insider trading laws, disclosure requirements, blackout periods, and surveillance systems are deeply embedded in securities regulation. Prediction markets — particularly decentralized, blockchain-based platforms — lack equivalent safeguards. The CFTC has jurisdiction over these platforms as derivatives markets, but the enforcement framework remains immature by comparison.
Spagnuolo’s case tests whether existing commodities law can effectively police a world of pseudonymous wallet addresses, cross-border blockchain transactions, and prediction contracts that blur the line between gambling and financial derivatives. The outcome will have implications not only for Polymarket but for the entire category of decentralized prediction platforms that have grown rapidly in popularity over the past two years.
For now, Spagnuolo is free on bond, his professional reputation in ruins, his future uncertain. The case serves as a stark reminder that access to internal data — no matter how routine it may seem inside a company — becomes criminal the moment it is used to gain an unfair edge in a market where other participants are betting blind.
