MARKETSBTC $63,067.00 -0.26%ETH $1,784.21 +0.64%BNB $588.44 +0.67%XRP $0.977150 -0.41%SOL $74.47 +0.65%DOGE $0.059012 +0.03%ADA $0.154156 +1.08%AVAX $7.08 +1.01%Prices are for reference only, not investment advice

Ex-Google engineer faces charges over Gemini 3.0 insider-trading bets

Ex-Google engineer Michele Spagnuolo could be facing a new charge for betting on the release date for his former employer’s Gemini 3.0 AI model. 

Spagnuolo’s lawyers disclosed the federal prosecutors’ plan to add the new charge to a Manhattan federal court, according to Inner City Press.

The AI model launch dates that prediction markets now run contracts on are the kind of information that is usually available to staff working at the AI labs building them. Which is why the possible charge Spagnuolo faces could be precedent-setting for a new class of insider trading targets.

Former Google engineer faces new insider trading charge

The new charge federal prosecutors plan to add to Spagnuolo’s case is unlike the existing one, which rests on Google search data. 

According to the defendant’s lawyers, citing a November 17 phone conversation, prosecutors signaled that they might supersede the indictment with a count tied to prediction market trades on when Google’s Gemini 3.0 flagship AI model would be released, as reported by Inner City Press.

The new model release date charge, if it gets tacked on top of Spagnuolo’s case, could change how federal prosecutors proceed with the case. On one hand, regular consumers can access search rankings data. On the other hand, information about when a model will be released is routinely available to most staff members inside AI labs. 

Investigators reportedly obtained a statement from Spagnuolo about where Google’s systems stored information on the Gemini 3.0 rollout, then reached out to Google for related records one day later.

What are the current charges against Spagnuolo? 

The U.S. Attorney’s Office for the Southern District of New York (SDNY) unsealed a complaint charging the 36-year-old Spagnuolo, who traded under the alias “AlphaRaccoon,” with commodities fraud, wire fraud, and money laundering. The underlying case against the Switzerland-resident Italian citizen opened on May 27, 2026.

Prosecutors claimed that Spagnuolo used an internal Google tool marked confidential, clearing about $1.2 million of the roughly $2.75 million he staked on Polymarket listings tied to his former employer between October 15 and December 4, 2025. 

U.S. Attorney Jay Clayton said, “Spagnuolo violated the duties he owed to his employer and used Google’s confidential business information to make more than $1.2 million in trading profits on Polymarket.” 

The case, USA v. Spagnuolo, sits before U.S. District Judge Analisa Torres. On October 23, she suppressed part of his statements, and the government is appealing that ruling to the Second Circuit. 

A December 21, 2026 court date will weigh whether other statements can be used as evidence. Trial on the existing charges is set for June 1, 2027.

Spagnuolo, who pleaded not guilty, has been placed on leave at Google. He was released on a $2.25 million bond.

The Commodity Futures Trading Commission (CFTC) filed a parallel civil complaint. 

Insider trading could become a problem with AI model launches 

As information about model releases is openly accessible to AI lab employees, prediction markets offering contracts on these launches could open a route to a new class of insider trading cases. 

Four Polymarket accounts reportedly bet, a week in advance, that OpenAI would ship a new model by December 13, 2025. OpenAI released GPT-5.2 on December 11, and the accounts booked more than $13,000. A separate account turned Google’s 2025 search data into over $1 million in a single day.

Earlier this year, OpenAI fired an employee for using confidential company information on prediction markets, including Polymarket. KPMG partner Conway Dodge said corporate questions about writing prediction markets into insider trading policies have at least doubled in six months.

Cryptopolitan has reported that two federal cases already test insider trading risk in prediction markets, and that Polymarket has leaned on on-chain forensics to flag suspicious wallets.

If you're reading this, you’re already ahead. Stay there with our newsletter.

Source: CryptopolitanOriginal URL: www.cryptopolitan.com