This post was drafted with AI from my event notes and approved by me.
I was privileged to follow Roshdi Ahmad’s keynote on cybersecurity at GovXcellence Summit 2026. The room was full of senior civil servants from across the public service, and I knew they would want something practical.
I began with a story about sustainable development and the first time I used ChatGPT, and how close I came to letting it do my assignment for me. I used that to explain how our two teams work. My side is policy and Hafidz’s side is innovation, and like that assignment, the two only work when they go together.
Then I put the script aside and walked through the governance tools we have today: the AI Governance and Ethics guidelines, MyAI Standards, the AI@Work materials, JDN’s public sector guidelines, the AI Code of Ethics, and the direction of the AI Governance Bill. To make it concrete, I described my own use of agentic AI to prepare briefing packs and work through a heavy inbox, and how I apply governance to myself first.
I closed with three governance questions every public servant can ask before using AI, and with Churchill’s line that perfection is the enemy of progress.
The questions were good. One was about what organisations should do while there is no AI law yet. My answer was that existing laws, such as the PDPA and the Computer Crimes Act, already apply, so there is no reason to wait. Another came from a speaker from Japan, who asked which country leads in digitalising its public service. I said that coming a little later than others lets us learn from the EU, Vietnam and Korea, and that getting it right matters more than getting it fast. On the Bill itself, its public consultation has since ended, on 31 July 2026.
Thank you to Traicon Events and Preetty Sharma. The plan is at https://ai.gov.my/aiactionplan2030.