The June deadline 1 for central government departments to have reviewed their AI adoption plans has come and gone, yet many organisations remain stuck in the planning phase.
What’s preventing them from turning these strategies into action?
At a recent event we discussed these challenges with senior government representatives. They all agreed: AI risks are getting more attention than the potential benefits. This blog looks at the 5 key challenges that were covered and offers practical steps needed to move AI from concept to implementation.
1. The cycle of hesitation
Risk-averse leadership is holding back adoption, with many leaders hesitant to fully commit to AI projects. This slows down the development of workforce skills and in turn delays the adoption of new technologies.
Act now: Leaders should introduce AI through flexible, dynamic frameworks that allow safe experimentation. This should be paired with continuous reassessment to meet user needs quickly. By doing so, leadership can make more confident and timely decisions on AI implementation.
2. Making the case for AI investment
Funding remains a major hurdle, with 67% of public sector organisations citing budget constraints 2 as a key obstacle. While technology is advancing rapidly, the slow-moving budget approval processes have made it hard to fund AI projects.
Act now: Organisations need to balance cost-saving measures with the willingness to experiment. A clear strategy that connects AI initiatives to both user benefits and departmental savings can help secure funding. Rather than just responding to external pressures, focusing on user needs could reduce costs in the long run and lead to more effective AI implementation.
3. Data dilemmas in AI
Data management is a challenge, especially with issues like data sharing, storage, and access when working with partners. Privacy concerns have led 52% of public sector organisations to delay AI investments.
Act now: Open-source platforms with a common government directive can provide the transparency needed to ease concerns. A centralised approach to data management across departments would accelerate AI implementation while ensuring proper data governance and security.
4. Modernising legacy with AI
Modernising legacy systems has been a priority in the public sector for a long time now. And 57% of organisations still see outdated IT infrastructure as a major barrier, limiting AI’s potential. But if you look at it a different way – AI actually has the potential to improve legacy systems without needing to fully replace them.
Act now: AI can improve the efficiency and performance of some legacy systems, saving time and costs. Instead of doing a complete overhaul, departments should consider where AI can add value. Take a phased approach and modernise incrementally to avoid overwhelming existing systems.
5. GenAI: a blessing in disguise?
Public sector organisations have been slow to adopt new technologies, with only a few making significant progress in AI adoption. Generative AI (GenAI) could be the push needed to drive adoption.
Act now: While traditional AI still holds value, the excitement around GenAI could spark faster adoption in the public sector. Departments should use this momentum to explore both generative and traditional AI to meet their goals more effectively.
Still a way to go
AI adoption in the public sector is still in the early stages, facing challenges like cultural resistance, high upfront costs, and data-sharing concerns. The key to moving forward is clear communication and well-defined business cases. The more understandable and tangible AI solutions are, the easier it becomes to secure support and investment.
At Made Tech, we’ve helped public sector organisations like the Met Office use AI to deliver better outcomes. By improving user feedback, we helped them make more accurate weather predictions. Learn more in our recent talk at the Think Data for Government 2024 conference.
- https://www.nao.org.uk/wp-content/uploads/2024/03/use-of-artificial-intelligence-in-government-summary.pdf ↩︎
- https://www.thinkdigitalpartners.com/news/2024/10/10/slow-uptake-of-ai-in-government-hindering-strategic-goals-research/ ↩︎
About the Author
Jim Stamp
Head of Data and AI at Made Tech
Jim has over two decades of experience working in and around software development, spanning many disciplines. He has led the way in establishing Made Tech’s data capability while also leading several major data platform projects.
