Prime Minister Datuk Seri Anwar Ibrahim recently launched the Malaysia Digital 2030 (MD2030) blueprint, envisioning Malaysia becoming an “AI Nation” by 2030. It invokes a “whole-of-nation” strategy and states that harnessing AI is no longer optional but an inevitability that needs to be managed so that it “works for people”.

How Do We Measure Success?

Malaysia is not short of plans with lofty visions. Where we often fall short is in evaluating our direction and re-strategizing where necessary. Part of the problem, as many have pointed out, lies in implementation. But another part begins at the planning stage itself, particularly in how success indicators are set.

MD2030 states that “success is not about how many programmes are announced. It is about whether people experience better services, businesses become more productive and Malaysia builds stronger digital capability”. Unfortunately, the nine high-level targets identified in the plan do not fully reflect this aspiration.

Seven out of the nine targets are based on performance in global rankings. There are at least three reasons why this is not helpful. First, one can climb up a ranking simply because others perform worse, not necessarily because it has achieved material improvements. Second, when a target ranking is not met, one can attempt to escape accountability by attributing it to the progress of others. Third, the metrics behind these rankings do not always capture on-the-ground realities and can also be “gamed”.

The remaining two targets are a 30% share of the digital economy to GDP by 2030, and an annual growth rate of 1.6% in the Malaysian Well-Being Index (MyWI). The former does not clarify which other sectors’ shares are expected to shrink correspondingly, which parts of the digital value chain will drive this growth, or how it reflects genuine productivity gains. Meanwhile, the latter is difficult to attribute directly to digital transformation, given the many other factors that influence well-being.

The more specific sub-targets under each of the seven thrusts in the plan appear to suffer from the same issue: vague, unmeasurable, (perhaps) unrealistic, tangential, or open-ended — the opposite of the SMART (Specific, Measurable, Achievable, Relevant, Time-bound) goals we are all familiar with.

What Is a Plan For?

This brings us to a bigger question: what is a national plan for? Ideally, a national plan should articulate a country’s grand strategy and action plans for a particular vision. It should offer clear direction to public agencies, businesses, and citizens — for example, what part of their operations should change, what new compliance requirements to expect, what funding windows might open, where to invest, and what lifestyle’s adjustment to prepare. It should tell organizations how to order their priorities, allocate their resources, and manage their risks. It should offer predictability and guide coordination and collaboration.

But little of this can be gathered from MD2030. The plan remains rather abstract despite pointing to generally desirable visions. This is regrettable, given that the plan promised to lay out how we move “from vision to execution, turning ambition into a more practical delivery plan for the next five years”.

What Can We Do, Then?

To be fair, so far, at the times this opinion piece is written, only the summary document of MD2030 is publicly accessible, so the fuller details may be more specific. Ideally, the full plan should have been made available at launch. But as an exercise, let’s assume the details are still being worked out, and ask: how can MD2030 be translated into SMART-er goals?

Let’s start with a specific context: Malaysia is increasingly urbanized. That alone poses myriad challenges in urban planning, housing, transportation, public health, waste management, and service access, to name a few. At the same time, Malaysia is also committed to achieving net-zero greenhouse gas emissions as early as 2050. Digital transformation can, in some ways, offer solutions to these challenges. But it requires high-quality, usable geospatial data that is easily accessible to various parties, across government agencies, businesses, academics, civil groups, and advocates — the “whole-of-nation” that MD2030 emphasizes. This is not yet the case in Malaysia. So what goal can be set, as an example?

Specific: Build an interoperable, open geospatial data platform providing highly demanded or valued non-sensitive geospatial datasets under an open license that can be used by anyone. Provision of the datasets should be mandated by law to guarantee enforceability and predictability for businesses, especially local SMEs such as startups, who may design their business models around the availability of this data.

Measurable: An open data bill mandating the provision of such data, with relevant guardrails to mitigate potential risks, should be tabled in Parliament, say, within the next three years, alongside the development of the necessary infrastructure.

Achievable: Drafting and tabling a government bill is within the executive’s control, even though whether the bill passes or requires amendment is less certain. Besides, precedents exist — the EU’s Open Data Directive, for instance — showing that such legislation is both achievable and workable in practice.

Relevant: Such a bill is relevant to ensuring continuous access to open geospatial data, which in turn can drive digital innovation in both technology and business models. Without a reliable data foundation, businesses have little certainty to invest in building on top of it.

Time-bound: Specifying when the bill will be tabled gives businesses and public agencies a fixed point to plan around, rather than an open-ended promise that can be quietly deprioritized. A clear timeline also promotes accountability. It allows progress, or the lack of it, to be tracked — the same standard MD2030 asks of itself when it says success should not be judged by announcements.

From Ambition to Action

While the above is just an example, it illustrates the kind of detail an action plan should ideally contain. It is not meant to diminish the ambition behind MD2030. But ambition alone will not get us there. What will do is the diligence to translate that ambition into targets that are specific, accountable, and grounded in the everyday realities. I hope that the fuller version of MD2030, when it is made available, reflects this level of detail. Because in our encounter with AI, we need to be SMART-er first.

Ashraf Shaharudin is a Postdoctoral Scholar at the Center for Technology, Strategy & Sustainability (CTSS), Asia School of Business (ASB), Kuala Lumpur. The views expressed are his own and do not necessarily reflect those of CTSS or ASB.