Air Gapped Intelligence: Sovereign AI for the Regulated Market the Cloud Cannot Reach
Mickai is a sovereign intelligence operating system that regulated businesses own and run inside their own walls. Built and live today, it runs entirely on the customer's own hardware, air gapped, with zero data egress.
The regulated market the public cloud cannot lawfully reach
Mickai is a sovereign intelligence operating system that regulated businesses own and run inside their own walls. It is built and live today, not a concept or a roadmap. It runs entirely on the customer's own hardware, on premises and air gapped, with zero data egress and no public cloud round trip. That single design decision, keeping the intelligence where the data already sits, is what makes us usable by the firms that most need advanced AI and are least able to send their data anywhere.
We built for a specific and growing constraint. Around 0.85 million UK businesses, roughly 15 percent of the total, legally cannot send their data to public cloud AI. Across the EU the figure is closer to 5 million. These are not laggards waiting to be convinced. They are banks, insurers, hospitals, defence suppliers and critical infrastructure operators, held in place by rules that were written on purpose and are tightening, not loosening.
We serve that market directly. A large cloud platform reading over the buyer's shoulder should note the shape of the opportunity as well, because a platform that adds a sovereign layer reaches, in one move, the regulated demand it cannot serve from a shared cloud today.
Why the wall exists
The constraint is legal, not cultural, and it comes from many directions at once. The PRA's model risk expectations under SS1/23 put the burden of explainability and control on the firm. UK GDPR governs special category data with a strictness that public cloud inference struggles to satisfy. The NHS Data Security and Protection Toolkit sets the bar for patient data. The EU AI Act classifies much of this work as high risk. ITAR and EAR govern defence and dual use technical data. The NIS Regulations cover essential services. The US CLOUD Act means that data in a US operated cloud can be reached by a foreign government regardless of where the servers sit.
Any one of these is enough to keep a regulated workload off the public cloud. Together they define a market that is structurally ours. We did not lobby it into existence and we do not need it to change. We built the system that operates cleanly inside it.
What air gapped intelligence actually means
We run about fifty specialist models, twenty five domain and twenty five operational, entirely on the customer's own hardware. Cross model routing sits under a deterministic arbiter, so the same input produces the same output. Reproducibility is not a nicety here. It is what lets a regulated firm defend a decision to a supervisor, a court or an auditor, because the process can be shown, repeated and explained rather than described as a black box.
The sovereign models do the work of the studios. We name them from Greek myth and keep their function serious. Nemesis handles fraud and AML. Plutus covers finance and FP&A. Tyche runs underwriting. Prometheus does forecasting. Iris takes customer service. Nomos handles compliance. Astraea covers legal. Panacea handles clinical work. Pythia delivers business intelligence. Aletheia performs audit. Vinis provides voice. The Agentic Marketing Team runs marketing operations. Trust Agent holds the perimeter, and we offer OAR as a Service for firms that want the audit substrate on its own.
None of this leaves the building. There is no round trip to a shared cloud, no third party inference call, no telemetry quietly leaving the estate. The intelligence lives next to the data and stays there.
The Open Audit Record
Trust in an air gapped system cannot rest on the vendor's word, so we removed the need to trust us at all. Every consequential action is signed under post quantum cryptography (FIPS 204 ML-DSA-65, with ML-KEM-768) and hash chained into a tamper evident, append only ledger. We call it the Open Audit Record.
Anyone can verify it offline, for decades, without trusting the vendor and without a network connection. An auditor, a regulator or an opposing party can take the record and check it independently. The signatures are chosen to hold as quantum computing matures, so a decision made and signed today remains verifiable long into the future. This is the part that turns a fast model into a defensible system of record.
Attestation across fielded units
For estates that run many units, we built Pantheon, a post quantum Layer 1 currently on testnet. It provides multi node attestation across fielded units with no central server. Each unit can prove its state to the others without any of them phoning home to a coordinator that could be compromised, subpoenaed or simply switched off. Sovereignty holds not only inside one box but across a fleet of them, which is what a multi site bank or a distributed defence programme actually needs.
The intellectual property position
We have 104 filed UK patent applications, covering roughly 2,340 claims across 13 invention families, owned by Mickai LTD, with named inventor Mickarle Sean Junior Wagstaff-Irons. These are filed, not granted. Filing establishes priority and builds a prior art moat around the sovereign architecture, the audit record and the routing that holds it together.
We are a UK company, Mickai LTD, Companies House 17166618, with Birmingham manufacturing secured. Micky Irons is founder and CEO. The company owns the stack outright, which matters to any counterparty assessing where the value sits and who controls it.
The dual buyer thesis
The sovereign AI market was roughly USD 40 billion in 2025 and is projected to reach about USD 148 billion by 2032. We approach it from two sides. First, we sell sovereign AI directly to regulated firms that the public cloud cannot lawfully reach. Second, we license the patented stack to the platforms that want to reach those firms and cannot do so from a shared cloud.
Our internal analysis maps 196 companies and 311 patent to company pairs as potential licensees, with names including Microsoft, AWS, NVIDIA, Google, Adobe and IBM. That is potential licensee sizing, a view of where our filed IP touches existing products. It is not a signed book and not an infringement claim. We are an ally to the AI majors, not an OpenAI killer. The platforms want the regulated market. We built the lawful way in. Those interests line up rather than collide.
Where this leaves a regulated buyer
If you operate under any of the regimes above, the question is not whether advanced AI is worth having. It is whether you can deploy it without breaking the rules that keep you licensed. We built Mickai so the answer is yes. The models run on your hardware. The data never leaves. Every consequential action is signed and independently verifiable for decades. The system is live now, the IP position is filed and owned, and the market it serves is defined by law rather than by fashion.
Our pre-seed round is opening soon, and we welcome inquiries from interested partners by email at micky@mickai.co.uk or on LinkedIn.
Is Mickai a product I can run today, or a roadmap?
It is built and live today. The system runs on your own hardware, on premises and air gapped, with about fifty specialist models under a deterministic arbiter. It is deployed, not promised.
How can an auditor trust the system without trusting the vendor?
Through the Open Audit Record. Every consequential action is signed under post quantum cryptography (FIPS 204 ML-DSA-65, with ML-KEM-768) and hash chained into an append only ledger that anyone can verify offline, for decades, with no network connection and no reliance on us.
What is the state of the patents?
There are 104 filed UK patent applications, roughly 2,340 claims across 13 invention families, owned by Mickai LTD. They are filed, which establishes priority and a prior art moat. They are not granted, and we do not describe them as such.