Sovereign AI for HR and Recruitment: Keep Candidate Data In-House

Sovereign AI for HR and recruitment means running the models that read CVs, shortlist candidates and support people decisions on hardware the employer owns and controls, so candidate and employee data is never sent to a third-party cloud. The employer holds the model, the data and the logs in one place, offline and provable. This keeps personal data, including special category data, inside the organisation and under UK GDPR control, while every automated step stays open to human oversight and clear record-keeping.

• CVs, application forms and HR records hold personal and special category data that should not be handed to an outside cloud model.
• Sovereign AI runs on the employer's own hardware, so candidate and employee data stays in-house, offline and under the organisation's control.
• Automated decisions about people need human oversight, fairness checks and clear records, not a black box.
• An Open Audit Record captures every action, which supports transparency, bias review and UK GDPR accountability.
• Mickai runs as a SIOS on your own machines and is in closed beta for selective design partners at mickai.co.uk/beta.

Why is candidate and employee data so sensitive?

Recruitment and HR sit on some of the most personal information an organisation holds. A single application can carry names, contact details, work history, references, and sometimes special category data such as health information, ethnicity gathered for monitoring, or details revealed in an interview. Under UK GDPR this data must be handled lawfully, kept only as long as it is needed, and protected from unnecessary sharing.

The moment a CV or an HR file is pasted into a public cloud model, control leaves the building. The organisation can no longer say for certain where the data went, who could see it, or whether it was retained to improve someone else's system. For people data, that uncertainty is a real problem.

What is the risk of using a third-party cloud model for hiring?

Sending CVs, interview notes and personnel records to an outside model creates exposure that is hard to reverse. The data may be logged, cached or processed in places the employer cannot inspect, and a breach or a change of terms is outside the employer's hands. For special category data the bar is higher still, and a candidate consenting to be screened is not the same as consenting to have their file shipped to a third party.

There is also a fairness dimension. If a model shortlists or ranks candidates, the employer needs to understand and defend how it reached that outcome. A closed external service that cannot be inspected makes it difficult to check for unlawful bias or to keep the records that transparency requires.

How does sovereign AI keep HR data in-house?

Sovereign AI keeps the model, the data and the logs together on hardware the employer owns. The system reads CVs, drafts job descriptions, summarises interviews and searches personnel records on the organisation's own machines, with no dependence on an outside cloud. Because it runs offline, data does not leave the network, and the employer can prove where every record sits.

This changes the compliance conversation. Instead of trusting a supplier's promise about a distant data centre, the organisation can point to its own hardware, its own storage and its own audit logs. Data residency stops being a question of contracts and becomes a matter of physical fact.

How does this support fairness and human oversight?

Automated decisions about people should never run unattended. Sovereign AI is built to assist the recruiter or HR professional, surfacing information and drafting outputs while a human makes the call. Every action the system takes is written to an Open Audit Record, so there is a clear, reviewable trail of what was done and why.

That record supports the checks fairness demands. Teams can review how candidates were screened, look for patterns that hint at bias, and show the reasoning behind a shortlist. Transparency and record-keeping stop being an afterthought and become part of how the system works.

How does Mickai approach sovereign AI for HR?

Mickai runs as a SIOS on the employer's own hardware: owned, offline and provable. It brings recruitment and people-operations work into one place, from screening applications to summarising records, while candidate and employee data stays inside the organisation. Every action is captured on the Open Audit Record, which supports the transparency and human oversight that people decisions require.

Mickai was founded by Micky Irons to give organisations modern AI without surrendering their most sensitive data, and the approach is backed by 104 filed UK patent applications, approximately 2,340 claims, owned by Mickai LTD. The Mickai closed beta is open now for a small number of selective design partners, and employers who want to keep hiring and HR data in-house can apply at mickai.co.uk/beta. Not every applicant is accepted.

Frequently asked questions

Can I use AI for recruitment without sending candidate data to the cloud?

Yes. Sovereign AI runs the models on hardware the employer owns, so CVs, applications and interview notes are processed in-house and never sent to a third-party cloud. The employer keeps control of the data and the logs at every step.

Does sovereign AI still let a human make the hiring decision?

Yes. The system is designed to assist rather than replace the recruiter. It surfaces information and drafts outputs, while a person reviews and decides, and every action is recorded on the Open Audit Record for oversight.

How does keeping HR data in-house help with UK GDPR?

Holding the model, data and logs on the employer's own hardware means personal and special category data stays inside the organisation, with a clear record of where it lives and how it was used. This supports the accountability, transparency and record-keeping that data protection expects. This is not legal advice.

What is the Open Audit Record?

The Open Audit Record is a running log of every action the system takes. For HR and recruitment it gives teams a reviewable trail that supports fairness checks, bias review and transparency about how candidates were handled.

How do I join the Mickai beta?

The Mickai closed beta is open to a small number of selective design partners. Employers can apply at mickai.co.uk/beta, and applications are reviewed rather than accepted automatically.