The San Francisco Police Department began using BriefCam, an artificial-intelligence video-analysis platform, in August under a three-year contract costing $198,000. The department is using federal grant money for the software as part of a new Regional Vehicle Interdiction Desk intended to address crimes involving vehicles, including carjackings. BriefCam can sift through thousands of hours of recorded video, monitor live camera feeds and alert police to people matching selected descriptions. Its maker says the technology can track people with accuracy approaching facial recognition. SFPD’s version does not include facial-recognition tools, according to the department’s description of its system. The platform was developed by an Israeli technology company founded by Hebrew University professors. Israeli authorities have also used BriefCam in a camera network in East Jerusalem, where the Ministry of Housing and Construction contracted for the technology. The software is used by law enforcement agencies in the United States and abroad, as well as at sites including airports and stadiums. Its deployment in San Francisco brings renewed attention to the reach of police video monitoring: even without identifying someone through facial recognition, the system can help locate and follow people across footage. Critics have argued that this kind of tracking can be as invasive as facial recognition. The contract connects the technology to a specific police unit focused on vehicle-related offenses, but the system’s capacity to search large amounts of footage makes its use and oversight consequential for people captured by cameras across the city.