How Are Grab and Sea Group Using AI to Build Asia’s Most Powerful Super Apps?
Grab and Sea Group are deploying artificial intelligence across every layer of their super-app ecosystems, from ride dispatch and ecommerce personalisation to credit scoring and fraud detection. Grab reported its first full-year net profit in history in FY 2025 at USD 200 million,[1] with revenue of USD 3.37 billion (up 20% year-on-year, 18% in constant currency)[1] and adjusted EBITDA of USD 500 million (up 60% year-on-year).[1] Sea Group posted FY 2025 revenue of USD 22.9 billion (up 36.4% year-on-year),[2] and Q3 2025 net income of USD 375 million (up 145% year-on-year).[3] Both companies credit AI as a primary driver of their improving unit economics.
These are not AI companies in the way that DeepSeek or OpenAI are AI companies. They are platform businesses that have embedded AI so deeply into their operations that it has become inseparable from their competitive advantage. The distinction matters: Grab and Sea Group demonstrate that the largest commercial returns from AI in Southeast Asia may come not from building foundation models, but from applying AI to massive existing user bases.
How Does Grab Use AI Across Its Super App?
More than 90% of all Grab mobility rides are dispatched using AI,[4] according to the company. This AI-powered dispatch system enables Grab to double its city footprint while simultaneously reducing operations headcount.[4] The technology is particularly powerful for expansion into non-capital cities: auto-adaptive marketplace technology enables scaling to secondary and tertiary cities without proportional cost increases. Grab reports that GMV in non-capital cities is growing over 2x faster than in capital cities[5] — a direct result of AI reducing the per-city marginal cost of operations.
Grab serves 57 million monthly transacting users (Q3 2026)[6] across 900-plus cities in 8 countries,[1] having crossed 20 billion cumulative rides and deliveries since its founding.[7] CEO Anthony Tan committed to a nine-week “generative AI sprint” across the organisation in 2024,[8] signalling that AI integration is now a top-to-bottom priority rather than a technology team initiative.
The company’s autonomous vehicle partnership with WeRide launched the first public autonomous vehicle shuttle service in Singapore,[9] positioning Grab at the frontier of autonomous mobility in Southeast Asia. Grab’s 2028 targets include over 20% CAGR in group revenues from 2025 and USD 1.5 billion in adjusted EBITDA[1] — a 3x increase from FY 2025[1] — with AI efficiency gains compounding across merchant tools, AI-driven promotions, and autonomous vehicle partnerships.
What Is Sea Group’s AI Strategy and How Does Sailor2 LLM Fit In?
Sea Group’s AI strategy centres on three interlocking advantages: a data flywheel from 400 million-plus annual active users[2] across Shopee (ecommerce), Garena (gaming), and Monee/SeaMoney (fintech); the Sailor2 LLM, a purpose-built language model trained on 400 billion Southeast Asian language tokens across 1B, 8B, and 20B parameter variants[10]; and Sea AI Lab (SAIL), whose pipeline-parallelism research, building on its “Zero Bubble Pipeline Parallelism” method, improves AI training throughput by up to 55% over standard methods.[11]
Shopee’s AI personalisation engine has lifted purchase conversion rates by 10% year-on-year[12] and drove over 70% year-on-year advertising revenue growth in Q3 2025.[12] These are not marginal improvements — a 10% lift in conversion rates across a platform processing billions of dollars in transactions translates to hundreds of millions in incremental revenue. The February 2026 Google strategic AI partnership commits Google’s engineering resources to building an agentic shopping prototype on Shopee,[13] signalling that even the world’s largest AI companies see Shopee’s data as uniquely valuable.
Sea’s Monee (formerly SeaMoney) had USD 9.2 billion of loans outstanding at end-2025, with 1.1% non-performing (90+ days),[2] despite serving a largely unbanked Southeast Asian population. AI-driven credit scoring is the core technology enabling this performance: by analysing Shopee purchasing behaviour, Garena gaming patterns, and payment history, Monee can assess creditworthiness for customers who have no traditional credit score. This is AI-powered financial inclusion at scale.
How Do These Super Apps Compare to Chinese AI Platforms?
Grab and Sea Group are at the beginning of the AI monetisation curve that Chinese platforms like Alibaba and ByteDance have already traversed. Alibaba’s Taobao and Tmall deployed AI-driven personalisation years ago; Ant Group’s Alipay processes billions of AI-scored financial transactions. The comparison suggests significant upside for Southeast Asian super apps as their AI capabilities mature.
The critical difference is data diversity. Chinese super apps operate primarily within China’s domestic market, while Grab and Sea Group operate across linguistically and culturally diverse markets — Indonesia, Thailand, Vietnam, the Philippines, Malaysia, Singapore, and beyond. Sea Group’s Sailor2 LLM, trained on 400 billion Southeast Asian language tokens,[10] is a direct response to this challenge: a single model that can serve customers across multiple languages and cultural contexts.
Founder Forrest Li’s October 2025 memo outlined a vision of 10x growth to a trillion-dollar market capitalisation,[14] explicitly powered by AI. Sea’s investment in World Labs, Fei-Fei Li’s spatial intelligence startup,[15] signals ambition beyond ecommerce into next-generation AI applications. For Grab, the autonomous vehicle partnership with WeRide and the generative AI sprint suggest a similar trajectory of deepening AI integration.
The combined impact of these two companies on Southeast Asia’s AI landscape is structural. Sea alone has more than 400 million annual active users,[2] Grab 57 million monthly transacting users (Q3 2026),[6] and together they operate across the region’s largest markets. Grab is also moving into AI infrastructure: in November 2026, SoftBank and Grab announced agreements for a Malaysia AI infrastructure initiative with PETROS and plans for US$1.85 billion in initial combined seed capital.[16] As AI-native consumer platforms outside China, they represent the region’s best test case for whether super-app data advantages can compound into lasting AI competitive moats.
For more on the competitive dynamics between Southeast Asian ecommerce platforms, see our analysis of Shopee vs TikTok Shop.
Read more: the trillion-dollar AI opportunity in Southeast Asia and the reality behind SEA’s $159B ecommerce market
Sources & Further Reading
- Grab Reports Fourth Quarter and 2025 Results with First Full Year Net Profit — Grab Holdings, 2026-02-12
- Sea Limited Reports Fourth Quarter and Full Year 2025 Results — Sea Limited, 2026-03-03
- Sea Limited Reports Third Quarter 2025 Results — Sea Limited, 2025-11-11
- GRAB Q4 2025 Earnings Call — AllMind AI, 2026-02-11
- Grab Q4 Earnings Call Highlights — Yahoo Finance, 2026-02-17
- Grab Reports Third Quarter 2026 Results, Raises Full-Year Guidance — Grab Holdings, 2026-11-03
- Grab Q3 2026 Earnings Prepared Remarks — Grab Holdings, 2026-11-03
- Grab CEO Anthony Tan on AI-first innovations with heart and purpose — Grab, 2025-05-02
- Grab Q3 2025 Earnings Prepared Remarks — Grab Holdings, 2025-11-04
- Sailor2: Sailing in South-East Asia with Inclusive Multilingual LLMs (arXiv:2502.12982) — arXiv / Sea AI Lab, 2025-02-18
- Pipeline Parallelism with Controllable Memory (arXiv:2405.15362) — arXiv / Sea AI Lab, 2024-05-24
- Sea Third Quarter 2025 Earnings Call Transcript — Sea Limited, 2025-11-11
- Sea and Google Deepen Partnership to Deliver AI-Powered Innovations for Consumers Globally — Sea Limited, 2026-02-19
- Can Shopee parent really become a trillion-dollar company? — Momentum Works, 2025-10-27
- AI pioneer Fei-Fei Li’s World Labs raises $1 billion in funding — The Star, 2026-02-19
- SoftBank Group and Grab Sign Joint Development Agreements to Develop AI Infrastructure in Malaysia — SoftBank Group, 2026-11-09
Further reading