Scotland is very much embracing an AI future; something that was reinforced with the recent launch of Scotland’s AI Strategy 2026–2031 and AI Scotland.
And while some organisations are just beginning to map their AI journey, others – like the trailblazers within the Scottish Government and various NHS boards – are already grappling with the complexities of integration. Acknowledging this spectrum is vital. AI readiness isn’t a binary on/off switch; it is a curve.
Data foundations and technical readiness
As our appreciation and understanding of AI matures, so too are use cases. From generative AI experimentation to agentic AI tools, the scope for innovation is vast. But success comes back time and time again to the data that any systems are fed.
Moving from isolated experimentation to the agentic era – where AI doesn’t just suggest text but executes multi-step workflows – demands a shift in focus. The primary bottleneck is no longer “dirty” data, but a critical maturity gap in enterprise data fabrics.
To address this, our team has been working alongside public sector leaders at every stage of this curve, ensuring that technical foundations and data sets are robust.
Agentic AI requires semantic search capabilities and vector databases that can interpret the context of Scottish public life – from local legislative nuances to specific regional data sets. Open source, publicly available datasets are also being heavily used within the public sector to shape strategy.
But when it comes to digesting this data, if your cloud architecture treats data as a static lake rather than a dynamic, queryable asset, your AI will remain stuck in the pilot phase. Our specialists are currently helping departments bridge this gap, transforming legacy data repositories into high-performance engines for automation.
From governance to Human-in-the-Loop
This hands-on approach to AI will become even more vital as AI becomes more embedded. In considering what this future looks like, of course, governance is vital; but we will increasingly see teams take a Human-in-the-Loop (HITL) approach. This relatively new strategy integrates human interaction, oversight, and expertise into the machine learning lifecycle, acting as a functional architectural requirement rather than just a policy.
Gradually, HITL and the cloud are coming together to create hybrid AI systems that combine the massive processing power and scalability of the cloud with human judgment and expertise. This presents a fundamental shift from AI as a black box to AI as a collaborative, governed asset which will assist in maintaining public trust.
Human expertise must remain the final arbiter for high-stakes decisions in health, social care, or justice, directly addressing ethical AI by baking accountability into the infrastructure itself.
What next?
Whether you are at the start of your journey or looking to evolve your generative AI pilots into autonomous agents, the path forward is one of collaboration. We are actively partnering with Scottish public sector bodies to turn these plans into reality, ensuring that the infrastructure sitting beneath any AI is as sophisticated as the goals seeing to be achieved.
While navigating and benchmarking progress on the maturity curve, public sector teams must consider how the current cloud foundation supports the move toward observability and agentic workflows. With the inevitability of autonomous agents, a hybrid cloud/AI approach offers a clear path for leaders to move beyond pilots into robust, dependable, and verified AI environments that will serve all of Scotland’s citizens.
CirrusHQ is exhibiting at FutureScot’s Public Sector AI 2026 on 21 May. Drop by our stand to chat with the team who can help guide you to prepare your cloud for AI. https://futurescot.com/futurescot-conferences/public-sector-ai/