A Google software engineer allegedly turned his employer’s search data into a personal ATM, using confidential information to win roughly $1.2 million on Polymarket. If the charges stick, it could reshape how regulators think about prediction markets entirely.

Michele Spagnuolo, a 36-year-old engineer based in Zurich who has worked at Google since 2014, was charged on May 27 in the Southern District of New York with commodities fraud, wire fraud, and money laundering. Prosecutors say he accessed non-public data from Google’s internal “Year in Search 2025” tool, then used it to place winning bets on Polymarket under the alias “AlphaRaccoon.”

The AlphaRaccoon playbook

Between mid-October and early December 2025, Spagnuolo allegedly placed at least 16 bets on Polymarket tied to Google search trends. The data he accessed was classified as “Google Confidential,” giving him a significant edge over every other participant on the platform.

His most notable win came from correctly predicting that D4vd, the singer whose real name is David Anthony Burke, would be the most-searched person on Google for 2025.

In total, Spagnuolo risked approximately $2.75 million across his positions and walked away with around $1.2 million in profits.

By December 2025, the pattern of suspiciously accurate bets on Google-related events had already raised red flags. Spagnuolo was arrested and released the same day on a $2.25 million bond. He faces one count under the Commodity Exchange Act, which carries a maximum penalty of 10 years in prison, and wire fraud charges that carry a maximum of 20 years.

Why this case matters beyond one engineer

This is one of the first federal prosecutions to explicitly apply traditional insider trading principles to a decentralized prediction market. The government isn’t treating Polymarket like some gray-area crypto toy. It’s treating bets placed on the platform as commodities transactions subject to the same laws that govern futures markets.

Polymarket itself appears to have taken the situation seriously. The platform updated its market integrity rules in response to the AlphaRaccoon trading activity, adding explicit prohibitions against placing trades based on non-public confidential information.

What this means for the prediction market landscape

If courts uphold these charges, the precedent could ripple outward in several directions. Platform operators may need to implement far more rigorous compliance frameworks, including know-your-customer procedures and surveillance systems designed to detect patterns of insider activity.

For traders, the calculus changes too. Anyone with access to material non-public information related to a Polymarket event, whether they work at Google, a government agency, or a sports league, now has a clear signal that the feds consider using that information to place bets a federal crime.