There’s no question that Artificial Intelligence (AI) is already having an impact on securing art collections, and museums of all sizes must devote time to exploring the technologies already on the market.
Today, even medium-sized museums often have well over one hundred cameras in operation. Obviously, it’s impossible to monitor all of them at the same time, and the infrequency of events in the museum environment can lead to complacent or inattentive guards—especially when the museum is closed. However, if there wasn’t a risk associated with a certain location then there wouldn’t be a camera in the first place.
With AI-enhanced cameras, human monitors can be alerted in the control room, in the gallery, or both, when an identified risk emerges. Technology like Art Sentry “sees” when a person—innocent or otherwise—moves too close to a precious object. Museum security consultants universally agree that this technology is a one of the best practices recommended during museum security vulnerability assessments. We’re still in the early stages of this sort of tech, and we expect that the coming years will see drastic improvement in an already-valuable system.
We anticipate that the next important step in AI cameras—one that is on our doorstep—is software that will allow museum surveillance systems to learn the patterns of visitor behavior and alert guard staff in advance of individuals who require additional monitoring as they traverse the institution. This will provide security teams with intel about possible risks in real time.
In an industry known for high attrition among gallery guards, “experienced” cameras will be a great help in enhancing security.
AI can also be useful in identifying risks. Products like ChatGPT, Perplexity, Gemini, and others can answer complex questions on the topic and provide suggested solutions for complicated problems. In fact, the more intricate and detailed the question, the better the response.
For instance, a museum security leader might ask AI, “How to prevent art theft in a small museum” along with particulars about, say, the crime rate in the neighborhood, past incidents, police response times, and other important factors.
We would strongly advise against providing any identifying information about the institution, however, because questions gathered by these AI chats are not private and could eventually end up being used in ways the user does not anticipate. Moreover, none of these products are perfect and all responses should be carefully considered. Still, these are valuable tools for brainstorming and typically produce concepts and ideas that might not immediately come to mind.
Museum security professionals will be questioned about initiatives that include individual tracking systems and facial recognition, and each must be prepared to defend the systems in the face of privacy concerns or laws (depending on your country) and the political climate in the community and among leadership.
Some museums are already using AI biometrics like facial recognition. To others, such things are anathema. But museum security is, and will always be, an ongoing balancing act between access and protection. AI can be an important partner in keeping that balance.