In June, students at San Diego State University (SDSU) discovered through their student newspaper that more than 1,300 AI-enabled cameras had been installed across campus over the previous two years, including more than 330 in residence halls. The cameras, made by Avigilon, are capable of facial recognition, licence plate reading, behaviour analysis and crowd density tracking. SDSU says the advanced AI features are switched off, and the system is there to keep students safe. However, students say they were never told what the cameras could do, and that knowing they could be watched in those ways, including in the buildings where they live, changes how the environment feels.
While the SDSU story is set in the United States, the underlying questions it raises are relevant to student accommodation providers worldwide. As AI-enabled and data-driven technology moves from the back office into buildings, and from buildings into residential settings, providers need to be clear and able to explain where the line lies between care, convenience, safety and surveillance.
AI as a student companion and coach
For a growing number of students, AI is already part of how they manage their own wellbeing, with or without their accommodation provider’s involvement. HEPI’s 2026 student generative AI survey found that around 15% of UK students are already using general-purpose AI tools for companionship, advice, or to address loneliness. It also found that 8% of students use counselling or therapy services provided solely by AI, and 4% use services where counselling or therapy is provided partly by AI.
This does not mean that student accommodation providers should rush to replace human support with chatbots. If anything, it points in the opposite direction. Students are already experimenting with these tools, so the sector needs to be clear about what AI can safely do, what it should not do, and how it connects students back to real people.
Some universities are responding by building or procuring purpose-designed AI wellbeing tools rather than leaving students to navigate general-purpose AI alone. In the US, Wayhaven is one useful model: Furman University, Butler University and the University of Houston all describe it as an AI wellbeing coach or chatbot that helps students connect with existing campus resources. Importantly, it is positioned as a complement to human support, not a substitute for clinical care. Flourish sits closer to the AI wellbeing-intervention end of the spectrum, with emerging trial evidence suggesting potential benefits for belonging, resilience and loneliness.
In UK student accommodation, the more visible examples are often digital resident experience and wellbeing infrastructure rather than AI wellbeing tools. Student Roost’s partnership with Kinetic on the StudentLife platformd is one example, best described as a wellbeing case-management system for frontline teams. The StudentLife platform gives Student Roost teams greater visibility into at-risk students, enabling earlier intervention and support and signposting via a centralised system.
In the building
Behind the scenes, AI is increasingly supporting the work of keeping a building running – and, indirectly, keeping residents well. AI and data-driven systems are being used or explored for predictive maintenance, demand and occupancy forecasting, and identifying patterns that may help staff respond earlier to emerging resident issues – all things that matter to a student’s day-to-day experience of home without ever needing to “watch” them directly.
There’s also another category of sensor technology worth noting, precisely because of how differently it’s designed compared to camera-based systems. Some sensor-based systems use air-quality and acoustic analysis to detect smoking/vaping, unusual noise levels, or aggression. These are explicitly marketed as not using cameras or recording audio, for use in precisely the privacy-sensitive spaces (bathrooms, bedrooms, corridors) where camera-based monitoring would be difficult to justify. The design principle of solving a problem without capturing an image is instructive for how ‘safe AI’ in residential settings could be approached more broadly.
When safety becomes surveillance
This is why the SDSU story has important implications beyond California. The issue was not simply that cameras existed. CCTV and access-control systems are already common in student accommodation, and many students expect them at entrances, in corridors, and in shared spaces. The issue was the step-change in capability – there were cameras that looked familiar but could potentially analyse, categorise, or search for people in new ways.
This is important because it serves as a reminder that student accommodation is first and foremost a student’s home, rather than just an asset to be secured. Even where advanced features are switched off, unclear communication can create anxiety, suspicion and a sense of being watched. In that context, trust depends not only on what a system is doing, but on what residents have been told it can do, who controls it, and what safeguards are in place.
The lesson is not “don’t use AI for safety”. It is that once a system becomes AI-enabled, the resident conversation and transparency measures need to change too – ideally before installation.
The boardroom test
For UK and European operators, governance is beginning to catch up. UK data protection law already requires surveillance to be lawful, necessary and proportionate. The ICO’s guidance on facial recognition is clear that organisations need to assess the lawful basis, document their justification and consider the impact on people’s rights. The EU AI Act goes further, with particular restrictions around biometric and emotion-recognition technologies in sensitive settings, including education.
For accommodation providers, these regulations highlight that before deploying AI-enabled or sensor-based technology in residential settings, boards and senior teams should be able to answer six basic questions: what it collects; why it is needed; who can access the data; how long it is kept; what human oversight exists; and how students are told in plain language.
The best uses of technology in student accommodation will be those that make support easier to access, buildings easier to run, and risks easier to address without making students feel monitored in their own homes. A wellbeing tool that signposts human help, a maintenance system that prevents a leak, or a sensor that detects a problem without identifying a person can build trust. A camera network with unclear capabilities can quickly have the opposite effect.
Getting AI right in student accommodation may look like a technical or compliance challenge, but the SDSU story shows that it is also a trust challenge. The providers that handle it well will be those that can show students not only that the technology works, but that it is proportionate, transparent and genuinely there to support them.
