Singapore Launches Institute to Build AI Skills Across Its Public Service
The government-wide programme combines mandatory foundations, role-based learning and agency workshops, with a target to equip more than 150,000 officers by the end of 2028.
Edited by Tyronne Panaino
Singapore's Ministry of Digital Development and Information and Civil Service College launched the Institute of Digital Government on September 18 to coordinate digital, data and artificial-intelligence skills across the public service. The official announcement sets an end-2028 goal of equipping more than 150,000 public officers with practical capabilities for their day-to-day work and public services.
The programme matters to officials at every level, from new entrants to senior leaders, and to residents who encounter government services shaped by those workflows. It also offers a concrete test of whether a large public employer can move beyond general AI awareness toward repeatable training, controlled experimentation and measured workplace adoption.
A common foundation before specialist work
All officers are required to complete annual foundation modules covering cyber security, data protection and AI literacy. The institute then separates further development into three pathways: Aware 101, Apply 201 and Build 301. They are designed for different levels of proficiency and include practical exercises rather than relying only on passive instruction.
The stated organisation-wide target is for every public officer to reach Apply 201 proficiency by 2028, meaning they should be able to use technology effectively in everyday and domain-specific work. Product thinking comes before tool selection in the programme: officers are expected to understand user needs and define the problem before deciding how technology could help.
The institute is also continuing hands-on digital development for public-service leaders. The ministry says nearly 150 senior leaders had already completed foundational training spanning product thinking, modernisation, resilience, data and AI by the time of the launch.
Agency workshops turn lessons into workflow tests
A second layer is built around agency-level AI Sprints. These workshops give teams dedicated time to frame workplace problems, prototype and test potential solutions, and use enterprise tools in the context of their own responsibilities. Pilot sprints running since May 2026 had reached more than 1,000 officers across government, according to the ministry, and the institute plans to extend them to more agencies.
That structure distinguishes the initiative from a one-off AI course. Foundation modules establish a shared baseline; the pathways divide learning by proficiency; and the sprints connect training to an actual process that an agency may want to redesign. Cross-agency learning communities are intended to spread useful products and practices after those experiments. One pilot community brings together officers in non-technical roles who are already working on AI-enabled workflow changes.
What the evidence does and does not establish
The evidence is a primary government announcement, so it establishes the institute's launch, programme design and declared targets. It does not yet show how many officers will complete each pathway, how proficiency will be assessed across different roles, which prototypes will become production services, or whether the programme improves service quality. No comparative evaluation or independent outcome data accompanied the launch.
Those limitations are important because participation, capability and safe deployment are different checkpoints. The next useful evidence will be completion data, examples of agency deployments, governance findings from real use, and measurable changes to service delivery rather than the size of the training target alone.
Status
Confirmed. The institute and its programme were announced by Singapore's digital-development ministry. Internal confidence is medium because the operating plan and scale claims currently come from that single official source and have not been independently evaluated.
Sources
Update note: Last reviewed 2026-09-23. We will revise this post when Singapore publishes participation, completion or service-outcome evidence.
Sources
Drafted with AI assistance from source briefs; reviewed for citation completeness and label accuracy.