Google Engineer Accused of Turning Secret Search Data Into $1.2 Million Polymarket Windfall

A Google engineer who allegedly transformed confidential company data into more than $1.2 million in betting profits has become the latest figure at the center of a legal battle over whether prediction markets should be treated like Wall Street.

Federal prosecutors in New York have charged Michele Spagnuolo, a longtime software engineer at Google, with commodities fraud, wire fraud and money laundering after accusing him of exploiting nonpublic search trend data to place winning wagers on the cryptocurrency-based prediction platform Polymarket.

According to prosecutors, Spagnuolo allegedly used access to Google's closely guarded annual "Year in Search" rankings to effectively bet on outcomes he already knew in advance — a strategy that, if proven, would make predicting the future considerably easier than the rest of the market was led to believe.

The case is one of the most significant legal tests yet involving prediction markets, a rapidly growing industry where users buy and sell contracts tied to future events ranging from elections and sporting results to pop culture trends and economic developments.

At the center of the investigation is a Polymarket account operating under the name "AlphaRaccoon." Prosecutors allege that Spagnuolo, who had worked at Google since around 2014, had access to an internal system containing confidential data used to prepare Google's highly publicized annual search rankings. Every year, the company releases its "Year in Search" report, showcasing the topics, people and events that generated the most attention among users worldwide.

The rankings attract substantial media coverage and advertising interest, making the results commercially valuable and tightly restricted inside the company. According to court filings, Spagnuolo allegedly accessed those rankings before publication and then placed a series of large wagers on Polymarket that corresponded directly with information contained in the confidential data.

The alleged strategy was not necessarily to predict who would finish first in Google's rankings, but often to bet aggressively against highly popular candidates whom internal data already showed would not take the top position.

One example cited by prosecutors involved wagers backing musician Kendrick Lamar as the year's most-searched person while simultaneously betting against other heavily favored names, including Donald Trump and Bianca Censori.

Because prediction markets operate through fluctuating probabilities based on public sentiment, betting against crowd favorites can become extraordinarily profitable when the crowd turns out to be wrong. Allegedly knowing the answer beforehand makes the exercise less prediction and more accounting.

Over roughly 25 separate wagers tied to Google's annual search rankings, prosecutors say the AlphaRaccoon account risked approximately $2.75 million. When Google officially released its Year in Search results in December 2025, the account reportedly generated around $1.2 million in profit.

The investigation claims that after the markets settled, millions of dollars worth of cryptocurrency flowed into wallets connected to the account. Authorities allege that some of the funds were subsequently routed through multiple crypto transactions and services designed to make tracing ownership more difficult.

Online users had already begun speculating about AlphaRaccoon's identity by that point. Discussions on social media platforms and cryptocurrency forums reportedly suggested that only someone with inside access to Google's data could have placed such consistently accurate bets.

Soon afterward, investigators say the account's public username disappeared, replaced by an anonymous string of characters.

Federal authorities claim blockchain records ultimately connected the betting activity to cryptocurrency wallets linked to Spagnuolo. Prosecutors say one of those wallets transferred funds to an account registered under his name using Italian identification documents.

The charges carry potentially severe consequences.

Commodities fraud — the charge prosecutors are using to address the alleged misuse of confidential information on a prediction market — can carry penalties similar to traditional insider trading cases. The government argues that although Polymarket is not a stock exchange, its contracts function similarly enough to financial instruments that existing fraud statutes should apply.

The wire fraud charge stems from allegations that Google’s proprietary information was exploited for personal gain, while the money laundering count relates to alleged attempts to obscure the origin of profits after the wagers paid out.

The case arrives just weeks after another high-profile prosecution involving Polymarket.

Earlier this year, prosecutors charged US Army Special Forces Master Sergeant Gannon Ken Van Dyke with allegedly using classified information about a military operation linked to Nicolás Maduro to place profitable bets on the platform. Prosecutors claim he turned approximately $33,000 in wagers into more than $400,000 in profit. Van Dyke has pleaded not guilty.

Together, the two cases are rapidly pushing prediction markets into uncharted legal territory.

For decades, insider trading laws focused primarily on stocks, bonds and conventional financial markets. But as platforms like Polymarket blur the line between betting, forecasting and investing, prosecutors appear increasingly willing to apply the same principles to information-based wagering.

The central question may ultimately be simple: if using secret corporate information to buy stock is illegal, should using the same information to win a prediction market bet be treated any differently?

A federal court will now have to decide whether the answer is yes — or whether prediction markets have quietly become the newest frontier for insider trading enforcement.

Written by Thomas Nussbaumer