Every graduate I spoke to this month told me they know how to use AI.
Almost none of them could show me.
This is the difference between AI literacy and AI fluency — and right now, across Southeast Asia, it is quietly deciding who gets shortlisted and who does not.
What is AI literacy?
AI literacy is awareness. You know what ChatGPT is. You have used it to draft an email or summarise a reading. You understand, broadly, that AI is changing the workplace. Most graduates leaving university in 2026 have AI literacy. It is almost impossible not to.
But awareness is a floor, not a ceiling. And employers have stopped hiring for the floor.
What is AI fluency?
AI fluency is application. It means you can prompt effectively, iterate quickly, and critically evaluate the output. It means you know when AI makes your work better and when it introduces errors you need to catch. It means AI is not a shortcut — it is a workflow.
Fluency is using AI to produce a first draft of a client proposal, then knowing exactly which assumptions to pressure-test before it goes out. Literacy is knowing that a tool like that exists.
The gap between the two is enormous. And most curricula in Southeast Asia are still teaching literacy — if they are teaching anything at all.
What employers are actually looking for
Hiring managers across Singapore, Manila, and Jakarta tell us the same thing: they are not expecting graduates to be AI engineers. They want people who can be productive alongside AI tools on day one.
That looks like:
- Writing prompts that produce usable outputs, not just decent ones
- Knowing which tasks to delegate to AI and which require human judgment
- Fact-checking and editing AI output before it reaches a client or manager
- Using AI-assisted tools in the specific platforms relevant to their role — whether that is data analysis, marketing, customer service, or operations
None of these require a computer science degree. All of them require deliberate practice. And that practice is not happening at scale in most institutions yet.
Where graduates are getting stuck
The graduates who struggle in AI-augmented workplaces tend to fall into two camps.
The first camp over-relies on AI output and submits work that lacks critical thinking, is factually shaky, or reads as generic. Employers notice this within the first two weeks.
The second camp avoids AI altogether, feeling that using it is somehow cheating or that they do not know it well enough. They move slower than peers who use AI effectively, and that gap widens fast.
Both camps share the same root problem: they were taught what AI is, but not how to work with it. That is a curriculum problem — not a student problem.
What you can do right now
If you are a graduate or student, here is a practical starting point:
- Pick one tool and go deep. It is better to be fluent in one AI workflow than dabbling in five. Start with the one most relevant to your target role.
- Practice prompting, not just using. The quality of AI output is almost entirely determined by the quality of your prompt. Treat prompting as a skill and practise it deliberately.
- Build a portfolio of AI-augmented work. Show employers specific examples of how you used AI to produce a result — and what you brought to the process that the AI could not.
- Learn to spot errors. AI is wrong regularly and confidently. The most valuable AI skill in any workplace is knowing when to trust the output and when to push back.
For educators and institutions
The window to get ahead of this is still open — but it is closing. Graduates who enter the market in 2027 and 2028 will be competing against peers who had structured AI fluency training built into their programmes. The institutions that move now will see it in their graduate employment rates within two to three years.
AI fluency is not a technical subject. It can be embedded into any discipline — from nursing to accounting to communications — through project-based learning and real-world application. The question is not whether to include it. It is how fast you can move.
The graduates who are getting hired in 2026 are not the ones who know AI exists. They are the ones who use it confidently, critically, and well.
That is the bar. And it is only going to rise.
