University Police completed installing more than 1,300 AI-enabled cameras in 2024, and spread them across classroom buildings, bookstores, dining areas, parking structures, gyms and residence halls where students sleep.
The photo came to light only after investigative journalism students at The Daily Aztec loosened up camera locations with a public records request.
Where the cameras went tells a lot about what the system is designed to keep an eye on. More than 330 of them point to student accommodation, which is about 28 percent of the entire network.
Huaxyacac, the largest first-year dormitory, houses 79 cameras in itself. Tenochca and Chapultepec have 36 and 33. Eighteen of the school’s 24 residential buildings are included in the location list and none are mentioned in the license agreements that students have to sign before moving in.
The most difficult problem is what the cameras can do, a question the school never gave anyone an answer to.
These are Avigilon units and the manufacturer advertises a long menu of artificial intelligence features on its website, among them facial recognition, license plate recognition, object and intrusion detection, behavioral analysis, crowd density analysis and audio detection. SDSU purchased hardware that is able to recognize who you are, read your plates and analyze how you move.
La Monica Everett-Haynes, the university’s associate vice president and chief communications officer, said students are informed about the cameras through the Guide to Community Living handbook and the campus housing website. None of the documents say a word about the AI sitting behind the lens.
This silence applies directly to the school’s own rules. CSU policy states that cameras are in public areas, defined as “an area open to public access, where there is no reasonable expectation of privacy” and that they are prohibited from being “pointed in areas where there is a reasonable expectation of privacy, nor shall they be directed at or zoomed into the windows of any private residential building, including residence halls.”
Over 330 cameras are now installed in the residence halls. The handbook limits the promise to “communal elevators and other common areas (e.g., lobbies, lounges, laundry rooms, hallways, dining facilities, etc.),” which is a generous reading considering the dorm consisted of 79 cameras.
Putting pressure on all this, Campus Police built the network closer to a glorified motion sensor. “The upgrade supports basic motion detection in restricted areas to help alert staff when activity occurs outside of business or class hours,” Amanda Stills, the department’s public information officer, wrote in an email to The Daily Aztec.
“To be clear, they are not used for behavioral tracking, profiling or facial recognition.” Stills said the university intentionally limits features out of privacy and campus expectations, and has no plans to purchase more AI capability.
The Avigilon contract language tells a different story, describing an intent that reaches far beyond maintenance and efficiency. A camera that can do facial recognition is still a camera that can run facial recognition, no matter what the policy says today about keeping this feature turned off.
Students don’t get a map of where any of them live and the department wants to keep it that way. When asked about posting the signs, Stills said that posting camera signs would put public safety at risk.
“The University does not currently use specific signage for camera locations and has no plans to add such signage,” they wrote, adding, “Cameras are widely present in public spaces and common work areas on and off campus.”
The situation is like continuous recording without any notification at the point of recording, which forces students to believe that they are always on camera and are not sure when the AI is studying them.
SDSU sits here in front of its own system. All CSU campuses run some form of CCTV, and only California State University, Northridge has joined SDSU in rolling out AI-powered cameras.
This trend reaches far beyond California. Michigan State hired a contractor to build a system designed to “detect barrier violations, track individuals walking around campus, calculate crowd sizes, and read vehicle license plates,” and products like ZeroEyes, Flock Safety, and Volt AI are arriving on campuses across the country.
SDSU already has hardware on the walls, far more capable than the school is allowed to use. Whether it will remain inactive or not depends entirely on a promise and promises about monitoring have a short shelf life.
<a href