Air-gapped AI: operate without network dependencies
Design isolated AI with transportable artifacts, integrity verification, controlled updates and local logging.
An air-gapped realization must be complete: weights, tokenizer, dependencies, data, policies, verifiers and recovery documentation.
A dependency downloaded on first boot or a model referenced by alias makes the deployment non-reconstructible.
The decisive test is rebuilding and executing the realization in an environment with no network access.
What the architecture must make explicit
Privacy and sovereignty are properties of the executed system: network flows, data, telemetry, dependencies, updates, backups and authority must share one controlled perimeter.
Move from a keyword to an inspectable engineering decision
Bound
Define network, organizational and legal boundaries for the realization.
Inventory
Identify models, APIs, telemetry, registries, caches and bootstrap dependencies.
Isolate
Replace or encapsulate dependencies that violate the selected perimeter.
Challenge
Run no-network, service-loss and complete-restoration tests.
A claim is only as strong as its binding
The gate must evaluate the exact output produced by the exact realization. Integrity, scientific validity and authority are related, but never interchangeable.
Questions that change the design
What does air-gapped ai: operate without network dependencies require?
An air-gapped realization must be complete: weights, tokenizer, dependencies, data, policies, verifiers and recovery documentation. Privacy and sovereignty are properties of the executed system: network flows, data, telemetry, dependencies, updates, backups and authority must share one controlled perimeter.
Which evidence should be retained?
At minimum: exact realization identity, inputs and outputs, protocol version, measurements, failure cases and the decision rule. The decisive test is rebuilding and executing the realization in an environment with no network access.
Does AMF prescribe a particular model?
No. AMF treats the model, runtime, data, tools and deterministic components as candidate materials. Their retention depends on measured contribution and declared constraints.