The public sector faces a massive dilemma: organisations are under pressure to become more efficient with artificial intelligence, while remaining wary of losing control of sensitive citizen data to global tech giants.

After a successful proof of concept, UNE is now making a clear and strategic choice to solve exactly that challenge.

Through an expanded partnership with Norwegian technology company TellusR, they are building a secure foundation for the case-handling support of the future.

Far too many public AI projects stall because organisations suddenly realise they are about to give up control of their own data.
Morten Krogh-Moe CEO, TellusR
UNE is showing that it is entirely possible to innovate without compromising security. That such a demanding, security-critical organisation chooses to build its AI foundation on our technology is a major recognition that digital sovereignty actually matters in practice.
Morten Krogh-Moe CEO, TellusR

At the heart of the new partnership is TellusR’s platform, AI Communications Hub (ACH).

Instead of connecting directly and uncritically to an external language model, the platform acts as a secure control layer behind UNE’s own firewalls. From there, UNE decides which data is sent where, which models are used for which tasks, and who has access.

People — not machines — make the decisions

For UNE it is a non-negotiable principle that the technology must only free up time as a tool, and never replace human or legal judgement.

We are introducing AI in selected areas as support in case handling — for example anonymisation, translation and summaries of decisions.
Henning Syversen Head of Process and Information Management, UNE

He is crystal clear about where the line is drawn in their systems.

It is important for us to be clear about the distinction between support in case handling and the decision itself. AI does not make decisions at UNE. The final assessment and ruling are made by people.
Henning Syversen Head of Process and Information Management, UNE

Over time the platform will let UNE build and scale its own tailored AI assistants for case-handling support across the organisation.

Saying no to a digital straitjacket

For UNE it has been essential to build an architecture that guarantees data sovereignty and avoids vendor lock-in.

The solution makes it possible to swap language models and infrastructure as the technology evolves, without rebuilding the whole system. At the same time it provides built-in traceability — an absolute requirement in public administration, since every AI-driven search and process must be auditable.

The result is that UNE gains the digital agency to steer its own AI tools, instead of depending on closed foreign ecosystems.

Efficiency in government can never mean selling out control of the data. We are now building an engine room that gives UNE the power to steer the technology, rather than the technology steering them. These are exactly the kind of customers Norwegian software companies want from the public sector.
Morten Krogh-Moe CEO, TellusR

He adds:

A partnership like this also pushes TellusR forward. Together we create concrete Norwegian alternatives and take ownership of the AI architecture. UNE is a strong example of a public customer the way Norwegian software companies want them. In this way we are jointly strengthening European digital sovereignty.
Morten Krogh-Moe CEO, TellusR

About the Immigration Appeals Board (UNE)

The Immigration Appeals Board (UNE) is Norway’s national appeals body for immigration and citizenship cases. As an independent administrative agency, UNE’s societal mission is to safeguard due process in appeals against UDI decisions, in strict accordance with Norwegian law and Norway’s international obligations.

About TellusR

TellusR is a Norwegian technology company that develops the central AI architecture for the public and private sectors. The company has built the technology behind today’s language models since it was called machine learning back in 2012. With its core product AI Communications Hub (ACH), organisations get an independent platform to adopt artificial intelligence safely and flexibly — with guaranteed data sovereignty, full control of their own data, and without being locked into global tech giants.