The Licensing Economics of a 104-Application Sovereign AI Patent Estate

Why patents sit at the centre of a sovereign AI thesis

Most AI companies treat intellectual property as paperwork filed after the product ships. At Mickai we treat it as load-bearing structure. Mickai is a sovereign AI operating system, an SIOS: artificial intelligence that regulated businesses own and run inside their own walls, on-premises and air-gapped, with every action written to a tamper-evident, post-quantum-signed audit record we call the OAR. That is built and live, not a concept. The patent estate is the legal expression of how it works.

Today Mickai LTD owns 104 filed UK patent applications carrying 2,340 claims, with Micky Irons named as inventor. These are filed, not granted, and we are precise about that distinction. Filing establishes a priority date and a documented prior-art position. It does not assert that anyone is infringing anything. What it does is convert engineering decisions into dated, defensible claims that frame the field around how sovereign, auditable AI is constructed and operated.

This article explains the licensing economics that sit on top of that estate, and why a map of 196 companies and 311 patent-company pairs is a sizing exercise, not a litigation threat.

What the 311 patent-company pairs actually represent

When you file across a broad surface of sovereign-AI methods, you can analyse which operating companies work in adjacent territory. We have done that mapping carefully. The result is 196 companies that we identify as potential licensees, expressed as 311 patent-company pairs. Names in that set include Microsoft, AWS, NVIDIA, Google, Adobe and IBM.

Read that correctly. This is potential-licensee sizing. It is the disciplined question every IP-holding company should answer: if our filed claims describe methods that the wider market will need as regulated AI matures, which firms operate where those methods apply, and against which specific applications? A patent-company pair is a structured way of saying one filed application has commercial relevance to one named company's product surface. It is not an accusation. It is a market map drawn in the language of the estate itself.

The economic point is straightforward. An enterprise value built only on a product roadmap is fragile. An enterprise value that also rests on a dated, broad, independently mappable IP position has a second leg to stand on. The 311 pairs are how we make that second leg legible to an investor, a partner or an acquirer.

Why regulated buyers make the estate matter

The patents would be academic without a market that needs what they protect. That market is large and underserved. Roughly 0.85 million UK businesses, about 15 percent, legally cannot send their data to public-cloud AI. In the EU the comparable figure is around 5 million. The drivers are not preferences, they are obligations: PRA SS2/21 model-risk expectations, UK GDPR special-category data, the NHS DSP Toolkit, the EU AI Act high-risk regime, ITAR and EAR export controls, the NIS Regulations and the long reach of the US CLOUD Act.

For these firms, the standard hyperscaler AI pattern of shipping data out to a shared environment is a compliance event, not a convenience. Sovereign AI answers that, and the market reflects it. Independent sizing puts sovereign AI at around USD 40 billion in 2025 rising to roughly USD 148 billion by 2032. The patent estate stakes out methods inside exactly that expansion: ownership, on-prem and air-gapped operation, and tamper-evident auditability.

Mickai is an ally, and the buyer base is dual

We are not positioning Mickai against the frontier labs. The thesis is dual-buyer. One buyer is the regulated enterprise that needs to own and run AI inside its own perimeter. The other is the platform or model provider that wants to reach regulated customers it cannot serve directly under current data rules. Mickai is the sovereign layer that makes both possible. That is precisely why a potential-licensee map populated by Microsoft, AWS, NVIDIA, Google, Adobe and IBM is coherent rather than adversarial. These are companies whose customers increasingly need sovereign delivery, and a licensing relationship is a route to it.

How the live product turns claims into revenue surfaces

Patents underwrite value, but the live system is what creates it. Mickai ships as Greek-named Studio modules, each a working capability for a regulated function. Nemesis covers fraud and AML. Plutus handles finance and FP&A. Tyche serves underwriting, Prometheus forecasting, Iris customer service, Nomos compliance, Astraea legal, Panacea clinical, Pythia business intelligence and Aletheia audit. Around them sit Trust Agent, the AMT agentic marketing team, the Vinis voice layer and OAR-as-a-Service.

Each Studio is both a product line and a licensing surface. The methods that let Nemesis run AML inside an air-gapped bank, or let Aletheia produce a post-quantum-signed audit trail, are the same methods the estate describes. Product revenue and licensing economics are two reads of one architecture, which is the efficient way to build a defensible AI company.

A dated momentum signal

As a third-party signal of momentum, in June 2026 Micky Irons was verified live at number 4 on the Crunchbase CB Rank for people, with the Mickai company profile placing in the top 1 to 2 percent globally. We cite that as a dated June 2026 data point, not a permanent claim. Mickai is a UK company with Birmingham manufacturing secured, and we are building to scale.

Working with us early

We are speaking with selected partners as we scale. This is an invitation to be early in a company where the product is live, the regulated demand is documented and the IP position is filed and mapped. It is an opportunity. If you want to look at the estate, the Studios or the licensing map in detail, write to me directly at micky@mickai.co.uk.

Micky Irons, founder and CEO of Mickai.

Frequently asked questions

Does Mickai hold granted patents or filed applications?

Mickai LTD owns 104 filed UK patent applications carrying 2,340 claims, with Micky Irons named as inventor. They are filed, not granted. Filing establishes a priority date and a documented prior-art position, which frames how sovereign, auditable AI is constructed and operated.

What are the 311 patent-company pairs?

They are a potential-licensee sizing exercise covering 196 companies, including Microsoft, AWS, NVIDIA, Google, Adobe and IBM. Each pair links one filed application to one company's relevant product surface. This is a market map, not an infringement claim against any company.

Who needs sovereign AI and how big is the market?

Roughly 0.85 million UK businesses, about 15 percent, plus around 5 million EU firms, legally cannot send data to public-cloud AI under rules such as PRA SS2/21, UK GDPR, the NHS DSP Toolkit, the EU AI Act, ITAR and EAR, the NIS Regulations and the US CLOUD Act. Independent sizing puts sovereign AI at about USD 40 billion in 2025 rising to roughly USD 148 billion by 2032.

Is Mickai competing with OpenAI and the major cloud providers?

No. Mickai is an ally on a dual-buyer thesis. One buyer is the regulated enterprise that must own and run AI inside its own perimeter. The other is the platform or model provider that wants to reach regulated customers it cannot serve directly under current data rules. Mickai is the sovereign layer that makes both possible, which is why major platforms appear as potential licensees rather than rivals.

How can a partner get involved?

We are speaking with selected partners as we scale. The product is live, the regulated demand is documented and the IP position is filed and mapped. It is an opportunity to be early. To review the estate, the Studios or the licensing map, contact Micky Irons at micky@mickai.co.uk.