Google sees ex-engineers indicted for alleged Iran transfers

What happened: Google trade secrets theft charges summarized

A trio of Silicon Valley engineers were indicted on suspicion of conspiring to steal trade secrets from Google and other companies, with prosecutors alleging a coordinated scheme to remove confidential materials. As reported by the San Francisco Chronicle, the indictment was returned on Thursday and centers on proprietary technology tied to major tech firms.

Fox Business identified the defendants as Samaneh Ghandali, Soroor Ghandali, and Mohammadjavad Khosravi, and reported that authorities allege transfers of sensitive data to Iran. The reports describe conduct that, if proven, would implicate core intellectual property protections around advanced computing and AI development.

Decrypt reported that federal prosecutors charged the three former Google engineers with trade secret theft and obstruction, and that the defendants face up to 10 years in prison if convicted. The charges are allegations at this stage, and the defendants are presumed innocent unless and until proven guilty in court.

Why it matters: AI, national security, and Google trade secrets

The case lands as law enforcement increasingly frames trade secret theft involving AI as both an economic integrity issue and a national security concern. Separate recent prosecutions have focused attention on how corporate R&D in AI can be targeted by actors linked to foreign jurisdictions, elevating the stakes for compliance and security programs across the sector.

“the first-ever conviction on AI-related economic espionage charges,” said Assistant Director Roman Rozhavsky of the FBI’s Counterintelligence and Espionage Division, describing a separate case that underscores how authorities are prioritizing the protection of U.S. AI innovation. The statement signals an enforcement posture that treats AI-centric intellectual property as strategically sensitive.

As reported by The Guardian, Google has said it maintains strict safeguards to protect its confidential commercial information and refers suspected theft to law enforcement. That positioning aligns with the public-private model authorities emphasize for protecting critical innovation.

Immediate impacts: arrests, charges, and corporate safeguards

The immediate outcome is a set of criminal charges that could carry significant prison exposure if the government proves its case; custody or arrest details were not specified in the cited reports. Given the indictment posture, additional procedural steps, such as arraignments and discovery, would be expected as the case advances.

Defense arguments in a related AI trade secrets trial challenged whether certain documents were sufficiently secret or consistently labeled to qualify as trade secrets, as reported by Bloomberg Law. That debate illustrates why companies must be able to demonstrate robust access controls and coherent confidentiality regimes when prosecuting or defending trade secret claims.

At the time of this writing, Alphabet Inc. (GOOGL) closed at 309.00, down 0.63% on the day, with pre-market indications at 307.85 (-0.37%), based on data from NasdaqGS. Market moves are contextual and do not indicate case outcomes or legal exposure.

Allegations and evidence: Tensor processor files and Iran transfers

In this indictment, authorities allege the removal of sensitive corporate information and transfers to Iran; public reporting so far has not detailed the specific file types involved, such as whether any Google Tensor processor design materials were implicated. The central question in court will be whether the materials qualify as trade secrets and whether the defendants knowingly misappropriated them.

As noted earlier, defense positions in a separate AI-related case tested the boundaries of what counts as a trade secret by pointing to access breadth and labeling consistency. In practice, prosecutors must establish secrecy and economic value tied to the contested assets, while companies will need records and controls that substantiate both.