Daily life in advanced societies requires constant interaction between the government and the governed. Housing, employment, education, healthcare, transportation, public safety, welfare, taxation – these are just some of the matters that the authorities manage on our behalf, and on which citizens need – and are entitled to – information, advice and instruction.
Prompt and efficient transactions in the public sphere are obviously desirable, but significant labour and real estate costs can be incurred when enquiries can only be handled by human officers in physical government offices. The public too can become frustrated by long waiting times and inconvenient office hours – a politically unwelcome situation.
A solution increasingly adopted by authorities around the world is the use of chatbots. These virtual assistants, when trained well on large enough data sets and capable of natural language, are available to answer queries and provide information 24/7. This greatly relieves the burden on government officers and budgets alike.
The pressures of the pandemic greatly accelerated the application of AI to multiple aspects of governance, going far beyond chatbots advising citizens on filling out forms.
Analytics are now being brought to bear on such issues as tax evasion, infrastructure maintenance, prioritizing child welfare payments or predicting outbreaks of disease. Well-designed algorithms can sift through vast quantities of data to deliver actionable insights, saving costs and improving the speed and quality of public services. It must be noted that the pace of adoption of AI by government agencies depends on several factors, including budget, reliance on legacy systems, motivation of the leadership and the availability of the necessary skills.
There is however no doubt that the public sector trend is towards the increasing use of AI technologies. There is a growing use of robotic process automation (RPA) to automate certain back-office functions at government agencies, delivering efficiencies, better quality services and greater job satisfaction among public servants. At the more strategic level, AI and machine learning are helping the process of policy decision-making. The principles are the same – aggregating and analysing mountains of data to help design better policies, deliver more accurate forecasts, make better decisions, and improve communication and engagement with stakeholders.
Another area where AI is helping to advance national agendas is the Smart City concept, where the objective is the sustainable digital transformation of a community. Southeast Asia’s smart city initiatives are linked to the ASEAN Smart Cities Network, established in 2018. The network currently has 26 member cities, including Singapore and several cities in Thailand.
As more and more areas of governance are handed over to digital resources, certain issues inevitably arise. It is crucial for the success of any public sector AI initiative to ensure trust, accountability, fairness and an absence of bias. It is increasingly recognised that some algorithms are guilty of bias, particularly if facial recognition is involved.
Indeed, the most concerning risks with advanced analytics in the public sector are those of bias and discrimination, which can disproportionally affect vulnerable sections of the community. Examples are not hard to find, and this could help explain why some public agencies may be hesitant to deploy advanced analytics at scale.
ASEAN has already developed region-wide master plans and strategies to promote and develop the digital economy involving AI, and many ASEAN member states have established and continue to develop national AI strategies and governance frameworks. Among these Singapore, with its AI Singapore (AISG), has the most advanced national policy designed to drive AI readiness and regulate its deployment. Specifically, in 2019, Singapore’s Infocomm and Media Development Authority (IMDA) launched the AI Ethics and Governance framework at the Model AI Governance Framework gathering in Davos. Thailand, by contrast, has its National AI Ethics Guidelines, which derive from Digital Thailand, the country’s 20-year national strategy for digital transformation which includes AI adoption.
The other ASEAN states are at different levels of AI maturity, but we can be sure that the adoption of AI technologies in our region will continue to transform government decision-making and enhance engagement with citizens.