Imagine a world where engaging a government office is as simple as ordering a pizza. I recall an entire afternoon lost to a driver's license mix-up—stacks of paperwork, long lines, and the agonizing multi-hour wait. If only there were a magic wand to simplify the process!

That magic wand might be AI. Gone are the days of endless queues and bureaucratic hurdles. AI doesn't just cut through red tape—it revolutionizes how we interact with government services, automating repetitive and manual tasks, accelerating document processing, and reimagining the entire experience for customers and administrative officials.  

Estonia’s e-residency program is a stellar example of modern government efficiency with AI. This rollout in 2014, has since supported the creation of over 25,000 new Estonian companies and generated €213 million over the past decade. Thanks to AI, registration processes that once took several weeks are now completed in as little as 3 hours, reducing 90% of time spent by e-residents.  

Revolutionizing Government Efficiency

AI has permeated the public sector administration globally, showing great potential to transform government systems through several crucial applications:

1. Task Automation: AI can automate routine processes like application processing and record management. For example, Dubai’s AI assistant Rammas provides 24/7 support for bill payments, application tracking, and job applications, ultimately handling over 1.8 million inquiries in 2023.

2. Smarter Policy and Decision making: During the 2019 climate protests, Belgium Government used Citizen Lab’s AI-powered crowdsourcing tool to analyze public input, helping create 15 climate action policies based on public feedback while boosting transparency and trust in government decisions.

3. Improving Citizen Services: IRCTC’s AskDisha 2.0 chatbot in India handles 150,000 daily inquiries, automating 80% of processes like train bookings, hence reducing wait times, and enhancing user satisfaction.

Moving beyond just AI  

With the meteoric rise of ChatGPT, governments today turn their attention towards Generative AI (Gen AI). This emerging technology presents a dual challenge: it must be regulated effectively while also being harnessed to revolutionize service delivery, operational efficiency, and citizen engagement. The potential impact is staggering – according to Boston Consulting Group, Gen AI could drive productivity improvements worth $1.75 trillion annually across all levels of government globally.  

Recognizing the transformative potential, governments are quickly mobilising their administrative units to obtain first mover advantage in this AI race. Australia's Federal Government has taken a proactive stance, mandating every agency to appoint a committee chair responsible for AI oversight and publicly disclose their AI strategy.  

A critical question emerges: should governments develop their own large language model? While LLMs are fundamental to tools like ChatGPT, they represent only one aspect of Gen AI's capabilities. However, several governments, particularly in Asia, see immense value in creating bespoke LLMs. Singapore, for instance, has initiated the development of SEA-LION, a family of open-sourced LLMs designed to capture the linguistic and cultural nuances of Southeast Asia, trained on data in 11 languages.

Homegrown versions of these tools would certainly help government agencies enhance current processes in a secured environment. In an increasingly interconnected world, having national LLMs is not just an option for boosting government efficiency, but a strategic imperative for safeguarding digital sovereignty.

‍