Oakland police investigate theft of $11.4 million in Nvidia goods
The Oakland Police Department in California is currently investigating a theft of Nvidia goods valued at $11.4 million. According to Bay Area News Group (Mercury News/East Bay Times), the shipment contained six pallets of Nvidia hardware (computer chips and related hardware) and was originally scheduled to be transported from Oakland to Fort Worth, Texas.
The case shows that the cargo was picked up as planned in Oakland on August 29, 2026, by a truck bearing a food transportation company’s logo, and the pickup process matched the arrangement. The truck then drove south and lost contact. Police tracking found that the truck made multiple stops along the way, with individual stops lasting as long as several hours, before finally losing trace in San Bernardino County, California. The victim reported the case to Oakland police on September 13 (police confirmed that they were investigating a major theft that occurred in East Oakland). As of now, police have not announced the arrest of any suspects, the investigation is still ongoing, and the specific stolen product models or details of the crime have not been disclosed.
Experts pointed out that these chips sell for nearly twice as much outside the United States as in the U.S. Olaf Groth, a professor at the Haas School of Business at the University of California, Berkeley, said that a chip that sells for about $550,000 in the U.S. can sell for $1 million or more abroad, creating substantial incentives for hijacking at weak security points; as the shipment size grows and the crime becomes more complex, it shows a trend toward team specialization. He also noted that these chips are national security assets, and their whereabouts are unknown (possibly flowing to China or other conflict regions). In 2025, police in Fremont and Union City investigated theft cases targeting Nvidia shipments, with losses exceeding $7.5 million.
Source: Public information (Mercury News/East Bay Times, ABC7, etc.)
ABAB AI Insight
The core of the case is the exposure of vulnerabilities in the physical transportation link of the AI hardware supply chain. As key bottleneck components for current data centers and AI compute power, Nvidia chips have unit shipment values that have reached the tens-of-millions-of-dollar range, making traditional cargo security mechanisms (label verification, GPS tracking, routine stops) unable to match the density of value. After the pickup was completed “as planned,” the shipment lost contact, and together with multi-hour stops and the break in San Bernardino County, this fits the characteristics of professional cargo theft: using a legitimate appearance (food transportation markings) to complete the handoff, then shedding tracking through segmented transfers or concealment.
Compared with multiple Nvidia cargo thefts in the Bay Area in 2025 totaling more than $7.5 million, the single loss of $11.4 million is larger, showing that the criminal target has shifted from dispersed small batches to high-value concentrated shipments. The mechanism Olaf Groth described, in which overseas prices are nearly double, stems directly from U.S. export controls and the global shortage of AI computing power: demand in restricted markets lifts black-market premiums, creating an arbitrage path of “low-cost acquisition in the U.S. — high-price monetization overseas.” This mechanism makes the expected收益 of physical theft significantly higher than traditional cargo crime, while also introducing a national security dimension (the chips may be used for computing capacity construction in restricted regions).
There is currently no evidence showing the specific chip models, whether they have left the country, or whether any particular gang is involved, and the police have not released details, which also limits further judgment. Overall, the case reflects a shift in risk within AI infrastructure from the “compute production” link to the “compute circulation” link: the production end is highly concentrated under Nvidia’s control, while the circulation end relies on dispersed third-party freight and warehousing, and security investment lags behind the growth in asset value, giving professional criminal groups a continuing incentive to target high-value pallets.
ABAB News · Cognitive Law
When asset density rises, physical circulation security costs must grow nonlinearly; otherwise, professional theft will shift from a marginal event to a structural arbitrage.
Export controls raise black-market premiums, making “low-price acquisition in the U.S. — high-price monetization overseas” an ongoing incentive for high-value chips rather than a one-time event.
When supply chain control is concentrated on the production side, risk spills over to the circulation side; the physical transfer of high-value AI hardware becomes a new security and geopolitical weak point.