Partners back AI but data, skills and security are challenges

Partners weigh in on Gartner’s outlook for government AI adoption.

Sovereign AI and AI agents are expected to shape public sector adoption of AI within the next two to five years, according to Gartner’s 2025 Hype Cycle for Government Services.

By 2028, 65 percent of governments worldwide will introduce some technological sovereignty requirements to improve independence and protect from foreign regulatory interference, the research firm predicts.

In Australia, initiatives are already taking shape, with Sovereign Australia AI, Maincode’s Matilda homegrown LLM and ResetData’s sovereign AI-Factory.

Gartner also predicts by 2029, 60 percent of government agencies globally will leverage AI agents to automate over half of the citizen transactional interactions, up from less than 10 percent in 2025.

Arinco is finding that government agencies and departments are shifting focus from simple chat or knowledge-based solutions toward autonomous agents capable of taking real action and delivering outcomes

“It’s a shift that will inevitably reshape how work is done within government and across the broader public sector,” said Clancy Whittle, AI, apps and data practice manager at Arinco

Prompt engineering will be critical to maximising the value of generative AI in government. However, partners are ahead of Gartner, with prompt engineering already front and centre, not just future-looking.

“It’s the foundation of effective AI solutions and one of the most sought-after skills right now,” Whittle said.

Machine customers — nonhuman economic actors like connected vehicles transacting directly with government — will grow from 3 billion today to 8 billion by 2030, according to Gartner.

Partners agree machine customers are a little further off, but “inevitable over the next few years as service channels become more automated and interconnected,” Whitte told CRN.

Partners say data sovereignty is more urgent

Data sovereignty is moving towards maturity in the next couple of years, according to Gartner’s predictions. On the ground, partners are already reporting strong interest in data sovereignty from government clients and they believe sovereign capabilities are critical to AI adoption.

“The next wave of adoption will depend on models trained on Australian data and designed to reflect our cultural, ethical and legal context. That’s what will make these solutions truly useful and trustworthy for government,” Whittle told CRN Australia.

Furthermore, Whittle believes that sovereign data considerations need to consider not just where data is stored, but who controls it, who audits it and how it flows into downstream systems.

“Alongside that, ethical and responsible AI is critical,” he said.

Security is another important consideration, with protection built in at every layer to protect against cyber threats, data leaks and adversarial attacks.

Finally, capability building can’t be overlooked, public sector workforces need the skills to use and oversee AI responsibly.

“Governments will need to ensure fairness, transparency and explainability in decision-making, while also safeguarding against misuse, bias and unintended consequences,” he ended.