Meeting that demand is fuelling a huge expansion of the sector, with cumulative investment in European colocation facilities forecast to reach €176 billion between 2026 and 2031.
To meet such demand, data centre operators are under pressure to scale quickly, with physical security as important a priority as cybersecurity.
Data centres need to combine physical and cybersecurity to effectively protect their assets, data, and reputation. Cyber threats remain a major concern, with the European Union Agency for Cybersecurity (ENISA) identifying ransomware as the most impactful cyber threat to the EU and reporting that state-aligned threat groups have intensified their activity against European organisations.
Physical threats can be just as disruptive. Unauthorised access, theft, vandalism and sabotage can put people, equipment and business continuity at risk. Extreme weather events add another pressure, with wildfires and heatwaves impacting operations and damaging equipment.
This makes visibility critical. Data centre teams need to know not only who is entering a facility, but what is happening around the perimeter, inside server rooms and across critical operational areas.
Traditional CCTV provides an important record of activity, but modern AI-powered video can do more than capture events after they happen. It can help teams identify unusual activity as it occurs and focus human attention on situations that require further investigation.
At the perimeter, combining visual and thermal imaging with intelligent analytics can help detect movement in darkness, poor visibility or challenging weather conditions. This can provide an early warning when someone approaches a restricted area, while AI-powered analytics can distinguish potentially significant activity from common sources of false alarms, such as wildlife or moving trees.
The same approach can extend to vehicles entering and moving around the site. Licence plate recognition can help identify vehicles of interest, while AI-powered analytics can provide additional information such as vehicle type, make, model, colour or direction of travel. Operators can then search footage based on these visual characteristics and track a vehicle across different cameras, helping establish where it came from, where it travelled to and whether it remained on site.
AI-powered search can take this a step further. Some video management systems (VMS) now support semantic search, allowing operators to describe what they are looking for using natural language rather than relying on predefined attributes. For example, an operator could search for “a person carrying a phone” and the system can identify relevant footage based on the context of the scene. Similarity search can also help match similar sightings of a person across multiple cameras, making it easier to trace movement across a large site or, where systems are connected, multiple locations. Together, these capabilities can significantly reduce the time needed to find relevant footage and help operators build a clearer picture of what happened during an incident.
Once someone reaches the building, video can add another layer to conventional access control. A valid access credential does not necessarily mean the person using it is authorised to enter. Advanced video analytics can help identify tailgating, where an unauthorised individual follows an authorised person through a controlled entrance, while visual verification can help operators investigate whether the person entering matches the credentials being presented. Two-way audio can also allow remote security teams to communicate with visitors, issue warnings or provide instructions without needing someone physically present at every access point.
Within the facility, comprehensive video coverage becomes particularly important around sensitive areas such as server rooms and maintenance spaces. AI-powered video can identify people or vehicles appearing in unexpected locations, movement outside normal operating hours or unusual behaviour around critical equipment. For example, someone lingering near a server rack without an expected reason to be there could trigger an alert to operators to investigate further, even if the individual has already passed through an authorised access point.
Bringing data feeds and alerts into a central system can also give operators a more complete view of events as they unfold, facilitating faster and more informed responses and, if needed, post-event investigations. High-quality footage and searchable video can provide valuable evidence after an incident, helping teams reconstruct what happened, identify people or vehicles of interest and support investigations.
Of course, the video system within a data centre can become a target in its own right. Cyber resilience therefore needs to form part of the decision-making process when selecting physical security technology. This is becoming particularly important as the EU Cyber Resilience Act raises the bar for cybersecurity in products with digital elements.
Looking for evidence of secure-by-design practices and recognised cybersecurity credentials, such as the German BSI IT Security Label, can give data centre leaders greater confidence that the technology protecting their facilities is not creating another vulnerability.
The future of data centre security will depend on infrastructure that works together across physical and cybersecurity. AI can help operators focus on the events that need their attention while continuously monitoring for unusual activity across the site.
Done well, this creates a more resilient, layered approach to protecting the facilities, services, businesses and people that increasingly depend on them.
Subscribe to The Security Briefing for monthly updates!