A super-app and a frontier AI lab rarely share a training stage. When they do, the question is not whether employees will learn prompt engineering. It is whether the programme scales beyond one city and whether it teaches people to interrogate the model rather than trust it.
Grab and OpenAI have launched a regional AI skills programme starting in Singapore. The initiative aims to build AI literacy across Southeast Asia, beginning with Grab's home market. The partnership signals that large platform companies now see AI fluency as infrastructure, not a perk for engineers alone.
Most corporate AI training stops at tool demos. Employees watch a webinar, try ChatGPT once and return to email. This programme, by virtue of OpenAI's involvement, will likely include access to enterprise APIs, structured curricula and use cases drawn from Grab's operations in ride-hailing, food delivery and fintech. The regional ambition matters because Southeast Asia's digital economy is fragmented by language, regulation and infrastructure. A programme that works in Singapore must adapt for Jakarta, Manila and Hanoi if it wants to deliver on the "regional" claim.
The risk is that the training becomes a vendor showcase. OpenAI has a commercial interest in expanding enterprise adoption. Grab has a reputational interest in being seen as an AI-forward employer. The programme will succeed only if it teaches critical evaluation alongside capability - how to spot hallucinations, when to override a recommendation and how to measure whether an AI intervention improved an outcome.
This is not the first time a technology provider has partnered with a platform company to run skills programmes. Google and Temasek funded digital skills initiatives across Southeast Asia in the late 2010s. Microsoft and LinkedIn offer AI training through their enterprise bundles. What distinguishes this effort is the focus on a single company's workforce as the pilot, with regional expansion as the stated goal rather than a vague aspiration.
If Grab treats this as a genuine capability build - tracking who completes the training, which teams apply it and what changes in their work - the programme could become a template for other platform companies. If it remains a press release with light follow-through, it will join the long list of corporate AI announcements that generate headlines but no measurable skill shift.
Three elements will determine whether this programme delivers lasting capability. First, does it include hands-on practice with real workflows, or is it lecture-based? Second, does it train managers to coach AI use, or does it assume employees will self-organise? Third, does it measure behaviour change, not just completion rates?
L&D teams in large organisations should ask whether their own AI training efforts meet those standards. A partnership with a frontier lab is not required. What is required is a curriculum that moves beyond tool tourism, a cohort model that embeds practice and a measurement framework that tracks application, not attendance. Grab and OpenAI have announced the programme. The test is whether it changes how people work, not just what they know.
For organisations building AI literacy at scale, the lesson is that regional ambition must be matched by local adaptation. A single training deck will not work across markets with different digital maturity, language preferences and regulatory environments. The companies that succeed will treat AI capability as a product, not a project - iterating on feedback, measuring outcomes and adjusting content as the technology and the workforce both evolve.
Sources:
.png)
The WorkFlow podcast is hosted by Steve Glaveski with a mission to help you unlock your potential to do more great work in far less time, whether you're working as part of a team or flying solo, and to set you up for a richer life.
To help you avoid stepping into these all too common pitfalls, we’ve reflected on our five years as an organization working on corporate innovation programs across the globe, and have prepared 100 DOs and DON’Ts.
