When the odds of a little-known singer topping Google’s most-searched list sat near zero, one Polymarket trader went all in. He already knew the answer.

Between October and December 2025, Michele Spagnuolo — a 36-year-old staff information security engineer at Google — used internal company data to place $2.7 million in bets on prediction markets, according to federal prosecutors. By the time Google published its “Year in Search” results on December 4, Spagnuolo’s account, operating under the name “AlphaRaccoon,” had produced roughly $1.2 million in profit.

The FBI traced the cryptocurrency payments back to accounts linked to Spagnuolo’s Italian identification card. He was arrested Wednesday in New York, charged with commodities fraud, wire fraud, and money laundering. He was released on $2.25 million bond without entering a plea, according to ABC News. An attorney for Spagnuolo was not immediately identified.

A Surgical Exploitation

The mechanics were straightforward. According to the criminal complaint filed in the Southern District of New York, Spagnuolo accessed Google’s internal “Year in Search” data — a rolling tally of what the world was searching — through a company tool available to all employees. He then used that non-public information to place bets on Polymarket, a cryptocurrency-based platform where users wager on real-world outcomes through event contracts.

The trades adapted as the data shifted. Spagnuolo initially bet that Kendrick Lamar, the 2025 Super Bowl halftime headliner, would top the year’s search trends. But internal data showed alt-pop singer D4vd surging. Spagnuolo repositioned, placing bets on D4vd when the market priced his chances near zero and wagering against names like Bianca Censori and President Donald Trump.

D4vd, whose legal name is David Burke, was charged last month with murdering 14-year-old Celeste Rivas Hernandez. The criminal case drove search traffic — and Spagnuolo knew the trend line before anyone outside Google did.

He also correctly called other search-market outcomes, including whether Zohran Mamdani would rank in the top five most-searched people and whether Squid Game would top searched TV shows, according to a parallel civil complaint from the Commodity Futures Trading Commission.

The Data Access Problem

One detail in the complaint should unsettle Google. The data Spagnuolo exploited wasn’t locked in a classified system requiring special clearance. It was marketing material, accessible through an internal tool that any of Google’s roughly 180,000 employees could use.

“The employee accessed our marketing material using a tool available to all employees, but using such confidential information to place bets is a serious breach of our policies,” a Google spokesperson said in a statement. The company placed Spagnuolo on leave and said it is cooperating with law enforcement.

Spagnuolo is the one who acted. But the architecture that made it possible — a massive workforce with broad access to data that moves markets — is the company’s design.

Insider Trading, Redrawn

This is insider trading, but not as the law has traditionally understood it. There are no stocks, no conventional securities. Polymarket offers event contracts governed by the Commodity Exchange Act. The novel legal question is whether prohibitions against trading on material non-public information extend to prediction market wagers — and prosecutors are asserting forcefully that they do.

US Attorney Jay Clayton framed the charges in deliberately familiar language: “Corporate insiders cannot use confidential business information to turn a profit in our markets,” he said Wednesday. “Insider trading compromises the integrity of our markets, and the American people want this greed-driven conduct investigated and prosecuted.”

The case is the second high-profile insider trading prosecution involving Polymarket in as many months. In April, Army Special Forces Master Sergeant Gannon Ken Van Dyke was charged with using classified information about a US military operation to capture Venezuelan President Nicolás Maduro, netting more than $400,000 in Polymarket profits.

Both cases share a common anatomy: someone with privileged access to information placing bets on a platform that, by design, cannot verify what its users know.

The Structural Flaw

Polymarket has positioned itself as a cooperative partner. A spokesperson said the platform “is the only prediction platform to date whose cooperation has led to insider trading charges in the United States,” adding that “blockchain trading is transparent, traceable, and bad actors leave footprints.”

The platform recently rewrote its rules to explicitly bar users from trading on contracts where they hold confidential information. But rules are the easy part. Detection requires either whistleblowers or the kind of post-hoc forensic blockchain analysis the FBI conducted here — after profits have already been collected.

The deeper problem is structural. Prediction markets assume a level information playing field. That assumption breaks when participants bring non-public data to the table — whether from Google’s search logs, classified military briefings, or corporate earnings reports — and the market has no mechanism for identifying who holds which cards.

Sources