The Missing AI Layer
The prompt names AI sovereignty, sovereign compute, local models, compliance, digital identity, and GGUF/local AI as a missing strand. The existing site has AI identity-intelligence and resilient-web pages, but it does not yet explain AI infrastructure as a national-readiness layer.
This module treats AI sovereignty as the ability to run, audit, adapt, and govern critical AI capabilities without being fully dependent on someone else's cloud, model, policy, export condition, or opaque identity stack.
Compute, Models, Identity, Compliance
- Compute: sovereign or allied capacity for inference, training, evaluation, and emergency workloads.
- Models: open, local, fine-tuned, or domain-specific systems that can be inspected and adapted.
- Identity: digital identity that enables service delivery without becoming universal surveillance infrastructure.
- Compliance: model cards, audit logs, evaluation, data handling, and human recourse.
- Local formats: small and portable model deployment for offline, edge, classified, or degraded-network contexts.
Why It Belongs On defence.is
AI sovereignty connects the book's nervous-system layer to practical defencetech. A country that cannot run important digital functions under pressure is easier to switch off. A country that cannot audit AI systems is easier to manipulate, overcharge, or lock into brittle dependencies.
Design Boundary
The page should avoid hype and procurement fantasy. The right question is not whether everything should be nationalised or local. It is which workloads require sovereignty, which can safely use allied or commercial infrastructure, and what evidence proves the distinction.