November 3, 2026, (Inside AI) — Grab Holdings plans to deepen its investment in artificial intelligence and expand support for drivers and merchant partners. The move follows the collapse of its $600 million deal to acquire Foodpanda's Taiwan business, according to Peter Oey, the company's chief financial officer.
The Singapore-based superapp had announced the acquisition in March, aiming to consolidate its position in Taiwan's food delivery market. That deal was blocked by Taiwan's Fair Trade Commission in December 2024, citing concerns about reduced competition. Grab's pivot to AI signals a strategic shift away from acquisition-led growth toward internal capability building.
Oey said the company sees AI as a lever to improve operational efficiency and customer experience. The investment will target areas such as demand forecasting, route optimization, and personalized recommendations for users and merchants. Grab also plans to use AI to enhance support tools for its driver-partner network, which numbers over 5 million across Southeast Asia.
Grab's AI Push Follows Regulatory Setback
The failed Taiwan deal underscores the challenges superapps face when expanding through mergers in tightly regulated markets. Taiwan's regulator argued that the combined entity would control a dominant share of the food delivery sector, potentially harming competition. Grab's decision to redirect capital toward AI reflects a broader industry trend: technology firms are increasingly viewing AI as a core differentiator rather than a supplementary feature.
Read: Credit Saison AI ROI
Grab has already integrated AI into several services. Its AI-powered features include a chatbot for customer support and a predictive tool for driver allocation. The company has not disclosed the exact amount of the new investment, but Oey indicated it would be significant and spread over multiple years.
"We are committed to investing in AI to better serve our drivers and merchants, and to drive sustainable growth," Oey said. "The Taiwan deal did not proceed, but our focus remains on innovating for our ecosystem."
The competitive landscape in Southeast Asia is intensifying. Rivals like Sea Group and GoTo are also ramping up AI investments. Sea's Shopee has used AI for logistics optimization, while GoTo has deployed AI in its payments and mobility services. Grab's ability to differentiate through AI will depend on execution and data access.
Analysts note that Grab's AI push could help offset the loss of the Taiwan deal by improving margins and user engagement. However, they caution that AI investments require significant upfront costs and may not yield immediate returns. Grab's shares have been volatile since the deal was blocked, reflecting investor uncertainty about its growth trajectory.
Grab's move also aligns with a regional push. Singapore's government has been actively promoting AI adoption through initiatives like the National AI Strategy, which aims to deploy AI in key sectors including transport and logistics. Grab, as a homegrown superapp, is well-positioned to benefit from such policies.
The company's AI investments will likely focus on three areas: enhancing the driver experience through better earnings predictability, improving merchant tools for inventory management and marketing, and streamlining customer interactions. Oey emphasized that drivers and merchants are central to Grab's ecosystem and that AI should empower them, not replace them.
Read: AI Reshape Indian IT Services
Grab's failure to acquire Foodpanda Taiwan is not its first regulatory hurdle. In 2021, it faced scrutiny over its merger with Uber's Southeast Asian business, which was approved with conditions. The company has since focused on organic growth and smaller acquisitions.
Looking ahead, Grab's AI strategy will be closely watched as a test case for how superapps can leverage AI to drive growth in emerging markets. The company has not provided a timeline for the new investments, but Oey suggested that pilot programs could begin in the coming months.
Shamil has a low tolerance for AI that only works in controlled demos. He covers agentic systems and generative AI with the skepticism of someone who has seen too many “revolutionary” tools quietly disappear six months after launch.
