- Livelihoods of millions of gig workers and micro merchants now rely on its platform, and increasingly its AI
- With 3 digital banking licenses, aims to lean into using its data to lend to unbanked and underbanked in region
The middle class in Southeast Asia is feeling the pinch. Wages are not keeping pace with costs, credit is tightening, and trust in institutions is fraying. In that context, Grab’s president, Alex Hungate, tells Julie Nestor, EVP, marketing communications Asia Pacific for MasterCard that the company has chosen to lean into affordability and inclusion, not simply protect margins, as it navigates its next phase of growth. The conversation happened at the MasterCard Inclusive Growth Summit in Kuala Lumpur on 25 Oct.
At stake is not just Grab’s own business where US$13 billion (RM54 billion) a year worth of services and goods flows, but the livelihoods of millions of gig workers and micro merchants who now rely on its platform — and increasingly, on its AI.
Leaning into affordability for a squeezed middle class
On the back of riots in Indonesia in late Aug over cost of living issues, Hungate does not sugar-coat the backdrop. “The economic uncertainty is very difficult. You know, the middle classes in Southeast Asia, our region, are definitely being squeezed,” he says, pointing to “a breakdown in trust” and bubbling social unrest.
Grab now accounts for about US$13 billion a year in earnings for its partners — drivers, food merchants, and small businesses on the platform. “By some third party estimates, if those partners couldn’t find earnings on our platform, unemployment in this region would be between one and 2% higher,” he notes. That scale forces a different mindset in a downturn: platform decisions quickly become macro-relevant.
Instead of lifting prices to defend unit economics as customers struggle with affordability, Grab did the opposite. “We actually decided that we would lean into pushing down prices and affordability on our services,” says Hungate. The logic: if prices hold back demand, partners’ earnings will fall; if prices are cut, volume might compensate.
“So rather than have the earnings on the platform start to decline in a difficult economic environment, we reduced prices,” he explains. “The number of rides and the number of food orders and grocery orders has just gone up enormously because of the lower pricing. And in fact, we’ve been able to increase the earnings that the partners get as a result.”
The second pressure point is credit. Hungate, a former conventional banker, is blunt: “In this kind of environment, of course, credit tightens up… and that’s most difficult for our partners.” Small businesses and gig workers are usually first to be cut off from lending just when they need it most.
“That’s the other thing that we can do,” he says. “Really lean into using our data to be able to lend to people that are either unbanked or underbanked in the region.” With three digital bank licences, Grab is betting that its data on partner behaviour can underwrite borrowers traditional banks ignore.
Hungate frames this as a continuation of what Grab did during Covid — working with governments so that drivers and merchants could stay plugged into the digital economy. The implicit message: if the platform wants to keep growing in tough times, it has to help its participants survive those same times.
Giving small merchants a virtual back office
Asked what else Grab can offer its partners, Hungate reaches for a giant killing metapho from biblical times. “We always kind of think about David versus Goliath,” he says. In this version, David is the small merchant who may have good products but no marketing team, no supply chain expertise, no cashflow adviser, and for some, no motivation coach (more on this later).
Grab’s job, as he tells it, is “to try to level the playing field”. Digital storefronts on the app, customer service tools, delivery logistics, and cashless payments all give micro and small merchants capabilities they could never build alone. “They no longer have to just use cash. They can use digital payments. So all of these things level the playing field.”
Access to capital remains the big differentiator. Most merchants on Grab are micro SMEs, which “account for about 85% of employment in most markets, absolutely critical to our society,” he notes. “Most of them can’t get bank loans, and that’s something that we feel that we can do and we can do it sustainably.”
In short, Grab is positioning itself less as a pure marketplace and more as a virtual back office for the region’s smallest employers: payments, distribution, and increasingly, finance. The political subtext is obvious. If that 85% segment falters, social pressure rises. If they can tap growth via platforms like Grab, some of that pressure is eased.
AI at scale, with guardrails and a human touch
AI sits at the heart of the next phase of this story, and Hungate is alert to the risks. “AI brings additional challenges,” he says. It could “result in much higher youth unemployment”, which in turn is “related to social instability”. The question he poses is stark: “What are we going to do to make sure AI is a force for good and not something that creates further disruption and tension?”
Grab already runs about 1,000 AI models and is the only Lighthouse partner for OpenAI in Southeast Asia, getting early access to foundation models which it then tunes for local use. One example: when it first tried to use GPT-4.0 for speech recognition to help visually impaired users, “it only worked 45% of the time for Southeast Asia, and that was in English” because of accents and local place names. After collecting 120,000 voice samples from Southeast Asians, accuracy rose to 92%. “That’s not too bad, actually,” he says with some understatement.
On the ground, AI is being pointed at partners’ earning power. A driver “copilot” helps them position themselves through the day, alerts them to road closures, and lets them report issues by voice so they are not distracted on the road. For merchants, a chatbot has become an all-in-one adviser.
Sharing a recent conversation he had with a merchant in Indonesia, Hungate recounts how the merchant started by asking the chatbot for advice on increasing sales. The bot then analysed nearby restaurants, search trends, and product gaps before suggesting new menu items. When the merchant replied, “Oh, I can’t cook that,” the chatbot shifted gears, offering cooking advice, order planning, and lead times based on predicted demand. Finally, the merchat asked it for encouragement — and it obliged. “It’s like AI with heart,” Hungate says.
Behind the scenes, generative AI is dealing with operations from 800 cities that Grab operates in, digesting an incredible 300 terabytes of data a day on traffic, supply, demand, and merchant capacity. Humans simply cannot cope with that scale. Still Hungate is clear that experimentation must be balanced with governance. Grab has an AI standards framework and what he calls ‘brand guardrails’, codifying promises to drivers, merchants, consumers, governments, and communities — and then “programming” those into its AI systems.
The results are sometimes counter-intuitive. When demand spikes — rain at rush hour on a Friday night — prices surge, as expected. But now, an upward red arrow appears next to the price, effectively asking the user: “Do you really want to buy this right now? It’s much more expensive than normal.” As Hungate puts it, “How many companies do that?”
For Grab, the answer is simple: a platform that wants to be the earnings engine of a squeezed middle class cannot afford to lose their trust, no matter how clever its AI becomes.
Article was generated using AI with an editor overseeing the published version.
