The Dual-Buyer Thesis: Selling to the Regulated While Building a Category a Hyperscaler Would Want to Own

The market the cloud cannot serve

There is a large class of organisation that is legally barred from running public-cloud AI. Not reluctant. Barred. A bank governed by PRA SS2/21 cannot hand its model risk to an opaque external service. A hospital working under the NHS DSP Toolkit cannot let special-category patient data leave its boundary. A defence supplier under ITAR and EAR cannot place controlled technical data on infrastructure exposed to the US CLOUD Act. Add UK GDPR special-category rules, the EU AI Act high-risk obligations, and the NIS Regulations, and the picture is clear. Roughly 0.85 million UK businesses, about 15 percent of the total, and around 5 million across the EU sit outside the reach of the public-cloud AI model.

This is not a niche. The sovereign AI market was around USD 40 billion in 2025 and is on a path to roughly USD 148 billion by 2032. These organisations are not waiting for permission. They are waiting for AI they can own.

What Mickai is

Mickai is a sovereign AI operating system, a SIOS. It is AI that a regulated business owns and runs inside its own walls, on-premise and air-gapped, with no dependency on a third-party cloud. Every action the system takes is written to a tamper-evident, post-quantum-signed audit record called the OAR. That is the part the regulator cares about. It is not enough to run AI privately. You have to prove what it did, in order, with cryptographic integrity that survives the arrival of quantum computing. This is built and live.

On top of that substrate sit the Studios, each a domain capability with a Greek name. Nemesis handles fraud and AML. Plutus runs finance. Tyche does underwriting. Prometheus forecasts. Iris carries customer service. Nomos covers compliance, Astraea legal, Panacea the clinical domain, Pythia business intelligence, and Aletheia audit. Around them sit Trust Agent, AMT, the Vinis voice layer, OAR-as-a-Service, and HELIOS hardware. The point is not the names. The point is that a regulated buyer can adopt a working capability inside their own walls without surrendering control of their data, their model, or their evidence trail.

The first buyer: the regulated organisation

The first buyer is the institution the public cloud cannot serve. For them the value is direct. They get production AI they own, that runs where their data already lives, and that produces the audit record their supervisor expects to see. The OAR turns a compliance liability into a compliance asset. Instead of explaining why an AI decision is defensible after the fact, the institution can hand over a signed, ordered, tamper-evident record of exactly what happened.

That is a clean commercial relationship. The buyer has a legal reason to choose sovereign infrastructure, a working product to deploy, and a margin profile that holds because the software does the heavy lifting. The Year 5 revenue path runs to billions at high gross margin, underwritten by demand that exists because of regulation rather than in spite of it.

The second buyer: a hyperscaler that would want the category

Here is the part that makes the thesis a dual one. While Mickai sells to the regulated market, it is building two assets that matter to the largest technology companies in the world.

The first is a patent estate. Mickai LTD holds 104 filed UK patent applications, roughly 2,340 claims, with Micky Irons as inventor. Filed, not granted, which is the right thing to say plainly. What filing secures is priority and a prior-art position. Sovereign, auditable, on-premise AI is exactly the territory the hyperscalers are moving into as their regulated customers ask for the same guarantees Mickai already ships. Mapping the estate against the field surfaces 196 companies and 311 patent-company pairs as potential licensees, including Microsoft, AWS, NVIDIA, Google, Adobe, and IBM. That is potential-licensee sizing, not signed revenue. It describes the shape of the moat. The companies most likely to need this ground are the ones who do not yet stand on it.

The second asset is the category itself. A hyperscaler cannot easily serve the cloud-barred market with a cloud product. The constraint is structural. To reach those customers a hyperscaler would need exactly what Mickai has built: a sovereign operating system, an audit substrate, the regulatory mapping, and the IP position underneath it all. That is the definition of a category a hyperscaler would want to own rather than rebuild.

I want to be precise about framing here. Mickai is an ally, not a challenger trying to unseat anyone. The hyperscalers are extraordinary at what they do. What they cannot do, by the nature of their model, is put their service inside an air-gapped bank or a defence prime. Mickai can, and does. That makes Mickai a complement to the cloud, and a complement with leverage.

Why the two buyers reinforce each other

The two buyers are not separate stories. They are the same story seen from two ends. Every regulated customer Mickai serves makes the patent estate more valuable, because each deployment is working proof that the claims describe real, shipping technology rather than theory. Every customer also deepens the category, the operational know-how, and the audit corpus that a future acquirer cannot simply assemble from scratch. Serving the regulated market is what builds the strategic asset. The strategic asset is what makes serving the regulated market durable.

This is why the dual-buyer thesis underwrites enterprise value. Revenue comes from institutions that must buy sovereign. Strategic value accrues to whoever ends up owning the category. Both grow from the same root.

Momentum and where we are

The signals are pointing the right way. As of June 2026, Micky Irons is ranked number four on Crunchbase, a third-party momentum measure, with the Mickai company profile placing in the top one to two percent globally. Mickai is a UK company with Birmingham manufacturing secured. We are building to scale and heading for the top.

None of that is the destination. It is evidence that the category is forming and that the market is finding us. The product is live, the estate is filed, the demand is structural, and the path is clear.

Micky Irons, founder and CEO of Mickai. micky@mickai.co.uk

FAQ

What makes Mickai different from cloud AI providers? Mickai is a sovereign AI operating system that runs entirely inside a customer's own walls, on-premise and air-gapped, with every action written to a post-quantum-signed, tamper-evident audit record. It serves organisations that are legally barred from public-cloud AI, which cloud-native providers cannot reach by design.

What is the dual-buyer thesis? The first buyer is the regulated institution that must own its AI for legal reasons. The second is a hyperscaler that would rationally want to own the category, because reaching the cloud-barred market requires exactly the sovereign substrate, audit layer, and patent position Mickai has built. Serving the first buyer builds the asset the second would value.

Are Mickai's patents granted? No. Mickai LTD holds 104 filed UK patent applications with roughly 2,340 claims, inventor Micky Irons. Filed, not granted. Filing secures priority and a prior-art position in sovereign, auditable, on-premise AI.

Who could license the patent estate? Mapping the estate against the field surfaces 196 companies and 311 patent-company pairs as potential licensees, including Microsoft, AWS, NVIDIA, Google, Adobe, and IBM. This is potential-licensee sizing, not signed revenue.

Frequently asked questions

What makes Mickai different from cloud AI providers?

Mickai is a sovereign AI operating system that runs entirely inside a customer's own walls, on-premise and air-gapped, with every action written to a post-quantum-signed, tamper-evident audit record. It serves organisations that are legally barred from public-cloud AI, which cloud-native providers cannot reach by design.

What is the dual-buyer thesis?

The first buyer is the regulated institution that must own its AI for legal reasons. The second is a hyperscaler that would rationally want to own the category, because reaching the cloud-barred market requires exactly the sovereign substrate, audit layer, and patent position Mickai has built. Serving the first buyer builds the asset the second would value.

Are Mickai's patents granted?

No. Mickai LTD holds 104 filed UK patent applications with roughly 2,340 claims, inventor Micky Irons. Filed, not granted. Filing secures priority and a prior-art position in sovereign, auditable, on-premise AI.

Who could license the patent estate?

Mapping the estate against the field surfaces 196 companies and 311 patent-company pairs as potential licensees, including Microsoft, AWS, NVIDIA, Google, Adobe, and IBM. This is potential-licensee sizing, not signed revenue.