On February 1 last year, Montana residents noticed a large white object in the sky, resembling another moon. This object turned out to be a Chinese spy balloon equipped with cameras and sensors, causing nationwide concern as it floated across the United States. The balloon’s surveillance potential was serious enough that a U.S. Air Force F-22 jet eventually shot it down off the South Carolina coast.

Simultaneously, approximately 30 cars owned by Chinese companies, equipped with cameras and geospatial mapping technology, were navigating the streets of Los Angeles, San Francisco, and San Jose. These self-driving cars were collecting detailed videos, audio recordings, and location data to enhance their autonomous driving algorithms.

Since 2017, Chinese-owned self-driving cars have covered 1.8 million miles in California, according to a Fortune analysis of the state’s Department of Motor Vehicles (DMV) data. These cars, as part of their functionality, capture detailed video and map roads with high precision. The collected data is then transferred to data centers for training the self-driving systems.

California’s state program allows companies, including Google-spinoff Waymo and Amazon-owned Zoox, to test autonomous vehicles on public roads. Among the 35 companies approved by the California DMV, seven are China-based, with five actively driving on California roads last year: WeRide, Apollo, AutoX, Pony.ai, and DiDi Research America. Some Chinese companies are also approved to test in Arizona and Texas.

Self-driving cars, fitted with cameras, microphones, and sophisticated sensors, have raised privacy concerns among advocates. Matthew Guariglia, a policy analyst at the Electronic Frontier Foundation, referred to them as “rolling surveillance devices” that collect extensive data on Americans.

Despite these concerns, the data collected by Chinese self-driving cars has received minimal scrutiny from U.S. authorities. Unlike Chinese-owned social media site TikTok, which faces potential bans due to national security concerns, Chinese self-driving cars have operated with little oversight. Some Chinese companies reportedly store U.S. data in China, making it accessible to the Chinese government. Security experts warn that such data could provide valuable intelligence for purposes ranging from mass surveillance to military planning.

Officials at state and federal agencies overseeing self-driving car testing acknowledge the lack of monitoring regarding the data collected by Chinese vehicles. Craig Singleton, director of the China program at the Foundation for Defense of Democracies, highlighted the absence of oversight, calling it “the wild, Wild West.”

Existing technologies like Google Maps and Street View already map much of the U.S., but they lack the updated and detailed data collected by self-driving cars. These cars, equipped with Lidar sensors, offer precise 3D scans of their surroundings. Such detailed mapping could be valuable for military and surveillance purposes.

In China, American-owned self-driving car companies face stringent regulations. They are not permitted to test their technologies in China, and high-precision maps necessary for self-driving are controlled by the Chinese government. In contrast, the U.S. lacks equivalent data security regulations for Chinese companies.

Despite the potential risks, some argue that autonomous cars are not the most efficient means of surveillance. Sam Abuelsamid, a car industry analyst, called concerns a “paranoid fever dream” but acknowledged the importance of questioning U.S. regulatory responses.

Chinese self-driving car companies have implemented measures to address privacy concerns. For instance, WeRide stated that its technology automatically edits videos to obscure human faces and license plates, deleting the original clips immediately after editing. However, the potential for these technologies to target individuals, like government or business leaders, remains a concern.

The regulatory gap in the U.S. regarding the data collected by Chinese self-driving cars contrasts sharply with the stringent measures in China, highlighting a critical area for policy development to address data security and national security concerns.