It fails at the boundary points teams overlook, especially USB ports, wireless radios, and third-party service workflows. Those paths can transfer code even when the core system never touches the internet. The practical test is whether every route into the environment is inspected, limited, and justified by operations.
Why Air Gap Protection Breaks at the Edges
air gap fail when teams define the boundary too narrowly. The core control may keep an OT system off the internet, but USB media, wireless interfaces, maintenance laptops, vendor access paths, and removable storage can still move code or data into the environment. That makes the real question not whether the asset is offline, but whether every ingress route is controlled.
In practice, the edge is where assumptions collapse. If a technician can plug in unmanaged media, if a radio remains enabled, or if a partner can push files through a support workflow, the air gap becomes a partial barrier rather than a true isolation control. The NIST SP 800-82 Rev 3 OT Security Guide treats segmentation and boundary control as operational controls, not a single design choice.
OT teams also need to distinguish isolation from trust. A device that never browses the web can still be exposed if it accepts signed updates, diagnostic files, or engineering downloads from another system that does have wider connectivity. That is why boundary review has to include the full transfer chain, not only the production controller or HMI itself.
What Boundary Paths Matter Most in OT
USB ports are the classic failure point because they bypass network controls entirely. Wireless radios, Bluetooth adapters, temporary hotspots, and cellular maintenance gear create the same problem by introducing alternate communications channels that security teams may not inventory as part of the OT boundary.
Third-party service workflows are just as important. Remote diagnostics, firmware staging, backup restoration, contractor file exchange, and shared jump hosts can all become delivery paths even when the plant network is nominally isolated. The right way to assess them is to ask whether the workflow can carry executable content, configuration change, or authentication material into the environment.
Asset visibility is the practical prerequisite for making that judgment. If you cannot confidently identify ports, radios, removable-media use, and vendor access paths, you cannot know whether the air gap is real or ceremonial. The OT Asset Visibility Guide is useful here because it focuses on discovery, inventory, and monitoring of exactly those boundary conditions.
How to Test Whether the Air Gap Still Holds
Start with an ingress map, not a topology diagram. List every way code, files, configuration, or credentials can reach the asset or its supporting tooling, then verify which of those routes are approved, inspected, and logged. If the answer depends on tribal knowledge, the control is weaker than it looks.
Then test the workflows that people actually use. In OT, the break often appears during maintenance, patching, backups, or vendor support, because those are the moments when exceptions are normalised. A protection model is credible only if it still holds when operations are under time pressure.
For many organisations, the most useful companion question is whether the environment has enough monitoring to notice a boundary violation before it becomes an incident. The CISA Industrial Control Systems resources are a good reference point for OT operational security practices and defensive validation.
Risk and Threat Considerations
Air gap failures usually do not begin with direct internet exposure. They begin with a trusted path that was never treated as an attack surface, such as removable media, unmanaged wireless, or a vendor workflow that can introduce executable content.
Failure mechanism: A boundary control is bypassed because the organisation secures the network perimeter but leaves alternate transfer mechanisms uninspected, unlogged, or permanently enabled.
Impact: Malware, malicious configuration changes, or unauthorised access can enter an OT environment even when the core asset remains offline, creating lateral spread, downtime, or process disruption.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
NIST CSF 2.0, NIST SP 800-53 Rev 5, CIS Controls v8 and CSA Cloud Controls Matrix set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| NIST CSF 2.0 | PR.PS-05 — Physical and Logical Access to Assets | Air-gap failure hinges on uncontrolled ingress paths into OT assets. |
| Recommendation — Control USB, wireless, and vendor access paths so only approved transfer routes reach OT assets. | ||
| NIST SP 800-53 Rev 5 | SC-7 — Boundary Protection | The question is about where isolation breaks at boundaries and alternate paths. |
| CM-7 — Least Functionality | Unused radios, ports, and services often create the failure points in air gaps. | |
| Recommendation — Enforce boundary protection at every transfer path, not only at the internet edge. Disable unnecessary ports, radios, and services that can carry content into OT systems. | ||
| CIS Controls v8 | CIS-12 — Network Infrastructure Management | OT air-gap failure is often a hidden connectivity and asset-visibility problem. |
| Recommendation — Inventory and manage OT connectivity points, including removable media and wireless interfaces. | ||
| CSA Cloud Controls Matrix | IVS — Infrastructure and Virtualization Security | OT isolation depends on secure segmentation and controlled connectivity boundaries. |
| Recommendation — Segment OT environments and validate every external connection path before allowing use. | ||
Practitioner Guidance
What to verify: Confirm that every OT ingress path has an owner, an approval rule, and an inspection point. If a route exists only for convenience, treat it as a security exception until it is justified operationally and technically constrained.
Common mistake: Treating “no internet access” as equivalent to “isolated.” In OT, the meaningful control is not internet absence, but whether all alternate delivery channels are controlled tightly enough to prevent code transfer.
Practitioner takeaway: Air gap protection fails where operations create unofficial bridges, so the practical test is whether the environment can resist real maintenance and vendor workflows, not only direct network reachability.