While much of the world has rushed to embrace AI, perhaps the most enthusiastic users are governments themselves.
Certainly, in the UK, the government has made clear not only its support for the sector as a route towards economic growth but has emphatically advocated for its use across public services.
As the public sector continues its journey of AI implementation, questions will arise over just what that could look like.
AI, it is worth remembering, is an extremely broad term that encompasses a vast array of technologies and tools, meaning implementation across different bodies can look drastically different.
In an effort to understand how public bodies are deploying AI, Dell Technologies and the International Data Corporation (IDC) have examined the trends among government decision-makers to determine what they see as valuable AI use.
Agentic AI dominates
The study found that among governments, there is massive interest in agentic AI, a category of tools that acts as a kind of artificial individual that can be given tasks and deployed autonomously.
According to the findings of the study, 71% of government decision-makers believe agentic AI is the best path to accelerate AI adoption in public services.
This is understandable and reflects the challenges faced by the public sector, notably workforce constraints, widening skills gaps and pressure to modernise services.
While these constraints encourage the assistance of agentic AI, there are fears from a majority (66%) that the technology is evolving faster than their workforce can keep up.
According to the study, more than half of leaders (51%) plan to invest in agentic AI over the next 12-18 months.
“Across the UK public sector, we are seeing strong interest in AI where there are clear, practical use cases that can deliver value for citizens and ease pressure on resources,” said Tariq Hussain, head of UK public sector at Dell Technologies.
Safeguarding
Though the findings suggest government leaders are bullish on deploying AI agents, for many, this is conditional.
Just under half of leaders (44%) said they will only accelerate AI adoption if strong safeguards covering data security and privacy are firmly in place.
This suggests an interesting divide among those willing to pursue adoption at all costs versus those maintaining a sense of caution.
“Leaders are taking a measured approach. The conversations we are having consistently focus on the need to get the foundations right first, from skills and governance to safeguards, security, and the infrastructure needed to support AI at scale,” added Hussain.
Another factor that has come up a lot among decision-makers is the capacity for sovereign data, that is, data which can be securely stored and accessed within the boundaries that it is from.
As many as 58% of government leaders identified strong sovereign data governance as a critical requirement.
