How One Public Sector AI Advisor Approaches Sovereignty To Turn Foreign Tech Into Domestic Value

The sovereign AI debate has moved past isolationism. In a global economy where every nation relies on imported GPUs, true digital independence isn't determined by what a country owns. Now, it's measured by what it creates. The new goal is turning foreign innovation into lasting domestic value, from job creation to public trust.

As Founder and CEO of Responsible Intelligence and an advisor to NATO and the UK Parliament, Maria Luciana Axente leads this shift in thinking. Today, she helps nations turn high-level strategy into practice with work that includes co-authoring PwC’s award-winning Responsible AI Toolkit and co-creating the Sovereign AI podcast. For Axente, the key to AI sovereignty is building rather than selling.

"Why sell access to that data when we can build companies on UK soil, keeping the value within our society?" Axente asks. "Patient data is both highly valuable and deeply sensitive." She points to the UK’s data pillar as the prime example.

The National Health Service holds a treasure trove of patient records, Axente explains. Currently, this attracts foreign tech companies seeking access through services like DARS. Yet a sovereign approach reframes this data as a strategic national asset used to build a domestic industry, rather than a commodity sold to the highest bidder.

To achieve this, Axente outlines a strategy based on three pillars: infrastructure, data, and skills. But beyond the economics, she says that sovereign AI serves a critical cultural function.

• Culture as code: Locally trained AI models offer a powerful tool for cultural preservation, Axente continues. "Can ChatGPT truly understand the cultural nuances of Wales or Essex?" she asks. "Can we expect it to crack a joke about Essex that only people from Essex would understand? I know a UK model is already being built to capture exactly those cultural flavors."

• Global game, local gain: Ultimately, nations must participate in the global tech ecosystem, including accepting foreign investment—even if the goal is simply to retain agency. "Countries must retain economic independence and agency," Axente explains.

Even with the right data and culture strategies, a domestic AI industry cannot exist without the third pillar: public consent. Research shows public trust remains a significant hurdle. To solve this, Axente says leaders must diagnose the root cause.

• Diagnosing distrust: The issue is rarely the government institution itself, according to Axente. Instead, the political layer above it is often to blame. "I have seen firsthand the stability a professional civil service brings. It is a godsend," she says. Unfortunately, that stability is frequently threatened by the disruptive agendas of changing political leadership.

Trust deficits deepen when politicians frame AI solely as a tool for "efficiency," Axente concludes. For the public sector workforce, that word sounds like a threat to job security. Rather than replacing the workforce, she says the path to rebuilding trust is to augment it. By automating low-stakes services, for instance, governments can free up public servants for high-touch interactions. "You don’t need to speak with a person about your bin collection. That can be automated. But when applying for benefits, you are at your most vulnerable and need a real human connection. The way to build trust is to recognize that human employees are the most valuable asset, use AI to offload repetitive tasks, and free up their time to focus on what matters: providing human-led services for people who are most in need."