Thirty thousand people. That’s the number of Grab partners — drivers, delivery riders, merchant-partners — that OpenAI and Grab are targeting with their new joint programme, GO Forward with AI. Announced on September 23, 2026, the initiative is one of the most concrete attempts yet to push practical AI skills in Southeast Asia beyond conference keynotes and into the hands of people who actually need them to earn a living. No research lab. No PhD required. Just working people learning how to use AI tools to do their jobs better.
Why Southeast Asia, Why Now
Southeast Asia is an interesting test case for AI adoption. The region has over 670 million people, high smartphone penetration, and a massive gig economy workforce. Grab alone operates across eight countries — Singapore, Malaysia, Indonesia, the Philippines, Thailand, Vietnam, Myanmar, and Cambodia — and its platform underpins daily life for tens of millions of people. Yet AI literacy across the region remains uneven at best.
This isn’t Grab’s first brush with technology-driven workforce development. The company has historically positioned itself not just as a ride-hailing and delivery app but as a broader financial and social infrastructure for Southeast Asia. That framing makes an AI skills programme feel like a natural extension — though whether that’s a genuine commitment or good PR is a fair question to ask.
For OpenAI, the timing tracks with a broader push to embed its tools in real-world, high-volume platforms. The company has been expanding its enterprise and partnership footprint aggressively. If you’ve been following OpenAI Academy’s recent expansion into structured learning paths, this programme fits the same pattern: OpenAI wants to be the de facto AI education layer for the world, not just the model provider.
What GO Forward with AI Actually Covers
The programme is built around practical, accessible training rather than theoretical AI concepts. That’s the right call. Telling a delivery driver in Jakarta about transformer architecture isn’t useful. Teaching them how to use AI-powered tools to manage their earnings, respond to customer queries faster, or optimise their routes — that’s something they can act on tomorrow.
Here’s what the programme is designed to deliver:
- AI literacy fundamentals — Basic understanding of what AI tools can and can’t do, framed around everyday work scenarios relevant to Grab partners
- Practical tool usage — Hands-on training with AI-powered features inside the Grab platform itself, as well as general-purpose tools
- Business and financial skills — Using AI to better manage income, track expenses, and plan for variable earnings typical of gig work
- Customer interaction — AI-assisted communication tools to help merchant-partners handle customer queries, complaints, and reviews more effectively
- Localised content — Training materials adapted for the linguistic and cultural diversity of Southeast Asia, which is non-trivial given the region spans dozens of languages
The 30,000 target figure covers a mix of driver-partners, delivery partners, and small merchant-partners who operate through the Grab platform. The programme is structured to be completed in digestible sessions, presumably via mobile — because if you’re designing something for gig workers, it has to work on a phone between shifts, not require a laptop and a quiet room.
OpenAI’s role appears to be primarily curriculum and content development, with Grab handling distribution and localisation across its network. That’s a sensible division of labour. OpenAI knows AI; Grab knows how to reach this specific population at scale.
How This Compares to Other AI Upskilling Efforts
This isn’t the only game in town. Google has been running Grow with Google programmes across Southeast Asia for years, with digital skills training that increasingly incorporates AI tools. Microsoft’s AI Skills initiative has similarly ambitious global targets. And various national governments in the region — Singapore’s SkillsFuture being the most developed — have their own AI literacy programmes.
What makes GO Forward with AI different, at least on paper, is the distribution channel. Grab has a direct, ongoing relationship with hundreds of thousands of gig workers. It’s not asking people to opt into a separate programme from a tech company they’ve never heard of — it’s embedding training into a platform those workers already use daily. That’s a meaningful structural advantage.
The 30,000 figure is modest compared to some of these broader initiatives, but focused programmes with genuine employer buy-in tend to outperform large-scale, opt-in digital skills courses that see high dropout rates. Quality of engagement matters more than headline numbers.
The Gig Economy Angle Deserves More Attention
Here’s the thing: gig workers are one of the groups most exposed to AI-driven disruption. Automated routing already shapes how much a driver earns. AI-powered demand forecasting changes when and where it makes sense to be on the road. Merchant-partners compete with algorithmically-optimised listings. These aren’t hypothetical future scenarios — they’re already happening inside Grab’s own platform.
Training workers to understand and use AI tools is, among other things, a way to reduce anxiety and build some degree of agency in a system that increasingly runs on algorithms they can’t see. That’s not a complete answer to the structural challenges of gig work, but it’s not nothing either.
I wouldn’t be surprised if part of the motivation here — beyond the genuine upskilling mission — is also about trust. Workers who feel the platform is investing in them are less likely to agitate for regulatory changes or jump to competitors. That’s a cynical read, but it’s probably not wrong to hold both things as true simultaneously.
What This Means for the Broader AI Skills Push in the Region
Southeast Asia is becoming an increasingly important front in the global AI race, and not just in terms of model deployment. Countries like Singapore, Indonesia, and Vietnam are making serious national bets on AI as an economic driver. The question of who controls AI education — and whose frameworks and tools become the default — is genuinely significant.
OpenAI embedding its tools and curriculum into Grab’s network means that for a substantial chunk of the Southeast Asian gig workforce, OpenAI’s products will be their first real exposure to AI in a professional context. That has long-term implications for platform loyalty, tool familiarity, and what workers expect from AI going forward.
There’s also a competitive dimension here. Anthropic, Google, and Meta are all pursuing partnerships and distribution deals across the region. Anthropic’s $2 billion partnership with Accenture is aimed at enterprise deployment, which is a different market segment — but the underlying dynamic is the same: whoever gets their tools embedded in the workflows people actually use will have a durable advantage.
Localisation Is the Hard Part
Southeast Asia is not a monolith. Bahasa Indonesia, Tagalog, Thai, Vietnamese, Malay, Burmese — the linguistic diversity alone is a significant engineering and content challenge. Add in the variation in digital literacy, device capabilities, and connectivity across urban and rural areas, and you start to appreciate why most global tech companies treat the region as an afterthought.
Grab’s local knowledge is genuinely valuable here. The company has spent over a decade learning how to build products that work across this diversity. Whether that translates into AI training content that actually lands with a delivery rider in Bandung or a small restaurant owner in Chiang Mai is something we’ll have to wait and see on. The programme’s success will ultimately be measured by behaviour change, not enrolment numbers.
For more context on how OpenAI is thinking about global education initiatives, it’s worth reading about OpenAI Academy’s expanded learning paths — GO Forward with AI feels like a regional application of the same strategic thinking.
Key Takeaways
- GO Forward with AI targets 30,000 Grab partners across Southeast Asia with practical AI skills training
- Content covers AI literacy, tool usage, financial management, and customer interaction — all mobile-first
- OpenAI handles curriculum; Grab handles distribution across eight countries
- The programme is notable for using an existing employment relationship as the distribution channel, rather than relying on opt-in
- Localisation across Southeast Asia’s linguistic diversity is the biggest execution challenge
- This fits OpenAI’s broader pattern of embedding itself in high-volume platforms to establish default AI tool familiarity
What is GO Forward with AI?
GO Forward with AI is a joint programme from OpenAI and Grab designed to deliver practical AI skills training to 30,000 Grab partners — including driver-partners, delivery riders, and merchant-partners — across Southeast Asia. OpenAI announced the partnership on September 23, 2026.
Who is this programme for?
It’s aimed at Grab’s partner network — primarily gig workers and small business owners operating through the Grab platform across eight Southeast Asian countries. It’s not a general public programme; you need to be a Grab partner to participate.
How does it compare to Google’s and Microsoft’s AI skills programmes?
Google and Microsoft’s regional AI skills initiatives are broader in scope but rely on users opting in through separate platforms. GO Forward with AI’s advantage is that it’s delivered through a platform Grab partners already use daily, which typically drives higher engagement and completion rates.
When is it available?
The programme was announced in September 2026. Rollout details across all eight Grab markets haven’t been fully specified, but given Grab’s existing infrastructure, regional availability is expected to follow relatively quickly after the initial launch phase.
Programmes like this one will be worth watching closely over the next 12 to 18 months — not for the announcement, but for the actual outcomes. If Grab can report meaningful changes in how its partners use AI tools and what effect that has on their earnings or business performance, it would be one of the more concrete data points we have on large-scale AI upskilling in a developing market context. And if it struggles with localisation or engagement, that’s equally instructive for anyone trying to solve the same problem.
