Engaging Stakeholders

The integration of artificial intelligence (AI) into public administration faces significant barriers, including legacy systems, institutional resistance, and ethical concerns. Legacy IT infrastructure – exemplified by the United States Inland Revenue Service’s reliance on a paper filing system designed in the 1960s, and Thames Water, the largest water company in the UK, still using 35-year-old Lotus Notes software – creates substantial implementation challenges. Successful modernisation requires strategic investment and phased approaches, as demonstrated by Estonia’s X-Road platform and Singapore’s Smart Nation initiative. Resistance to AI adoption stems from employee apprehension and public scepticism, necessitating transparent communication and stakeholder engagement. Critical failures like the UK’s A-level grading algorithm, the Dutch childcare benefits scandal, and Australia’s Robodebt scheme highlight the importance of robust contingency planning, human oversight, and ethical frameworks. Cybersecurity faces both opportunities and risks; AI can enhance threat detection capabilities yet enable more sophisticated attacks. Public trust remains fragile; according to a recent study by the Alan Turing Institute, 72 per cent of UK citizens require stronger AI regulations to feel comfortable with the new technology, while two-thirds report experiencing AI-related harms. Successful AI adoption requires dedicated funding, cross-agency coordination, algorithmic accountability frameworks, and hybrid human–AI systems that balance innovation with democratic values and human dignity.