Southeast Asia's Path from AI Momentum to Value

5 Strategic Moves for Enterprises to Leapfrog and Capture 5x AI Value. BCG | Google Cloud.

Opportunity

Agentic AI can unlock more than US$350 billion of enterprise potential across Southeast Asia — more than 9% of GDP.

The flags of the six ASEAN-6 countries are shown: Malaysia, Thailand, Singapore, Indonesia, Vietnam and the Philippines.

Despite leading enterprises in the region pacing forward, many SEA enterprises are facing adoption paralysis, with overall AI maturity trailing global peers.

Waiting may feel safe now, but the landscape is shifting fast – Future-built enterprises have already unlocked 5x more value from AI in 2025.

The path to 5x value is charted by 5 strategic moves in AI leadership, organizational commitment, scaled transformation, agentic architecture, and human-AI teams.

SEA's next leapfrog moment is here - enterprises win by pairing sharp strategic focus with AI at scale to achieve transformative impact.

1. Represented by the scale of work that could be augmented by AI, estimated by applying role-level augmentation rates to country-specific labor value

2. Combined GDP of ASEAN-6 countries in 2025

Overall maturity

Many SEA enterprises are still in the waiting room due to adoption paralysis, even as AI leaders chart their path forward. AI maturity score (0-100): SEA average 41, global average 48, Future-built enterprises 76+, and SEA leaders at 85 (SEA leaders are the Future-built enterprises in SEA).

Enterprises by stage: 16% Stagnating (AI is absent or ad-hoc), 46% Emerging (AI is experimental and local in scope), 28% Scaling (AI is embedded in core workflows), 10% Future-built (AI is a key source of advantage).

To many, waiting feels safe: frontier tech changes too quickly, outpacing adoption; stakes are high and mistakes are costly; true market signals are blurred by noise.

…but the ground is already shifting underneath.

Those who wait risk getting caught off-guard by competitors who are leapfrogging ahead at any moment.

Dimensions analyzed

Current strengths center on Customer Experience and other GenAI-driven capabilities, next step is to invest in agentic capabilities.

Dimensions where SEA enterprises outperform global peers on average: Strategy — AI ambition exists for selected domains, with value tracking in place. Customer Experience — Automated some contact-center channels and piloted predictive personalization. Operations — Basic tools implemented; some manual effort needed but consistency is improving. People — Selected roles use AI in tasks, ad-hoc training and upskilling are in place.

Dimensions where SEA enterprises trail global peers on average: Innovation — AI is used by some teams in production but yet to redesign workflows. Technology — Enterprise AI platform exists but yet to support dynamic orchestration and risk controls. Data — Data is reusable but not yet fully embedded. AI-supported decisioning is not yet in place. Op Model — Enterprise AI leader and select cross-functional teams exist, but roles are unclear.

Value stories

Despite shifting landscapes, leading enterprises in SEA have gained momentum in realizing value from agentic AI.

1 of 3 — Customer Experience. "Leveraging data and insights to drive rapid personalization." A large Asian telco reimagined its postpaid product sales motion by infusing agentic AI right from capturing the right data signals to enabling both traditional (call center, retail stores) and digital channels (app, chat, voice) to drive improved conversion. Quote: "Agentic AI helped us provide the right data-backed nudges to our sales team with the right product recommendations that our customers valued. We are moving very fast with these new capabilities at our fingertips" — Chief Consumer Officer. Result: +80% uplift in upsell.

2 of 3 — Operations. "Applying Agentic AI to drive procurement savings." A large Asian industrial manufacturer reimagined its procure-to-pay process by infusing agentic AI to detect unfavorable historical price variances and enable forward-looking spend transparency and prioritized savings opportunities across its category portfolio. Quote: "We picked a use-case that was messy from a tech and data perspective, but allowed us to get real PL impact. We are now looking to scale this horizontally across the group and compound on our initial wins - this is where AI really shines" — Chief Executive Officer. Result: US$ 50m in savings opportunities.

3 of 3 — Technology. "Transforming finance with AI workloads on an agentic platform." A large Malaysia-based bank built its AI ecosystem on a hybrid environment that leverages a cloud platform as the integration layer to connect legacy systems. The bank then deployed a diverse suite of AI agents on this platform, running the models as a private instance to securely deploy while maintaining strict governance and data protection in a highly regulated industry. Quote: "We've reached a new level of scalability and efficiency by eliminating infrastructure management and seamlessly integrating with our core systems. This allows us to dedicate our full focus to delivering smarter, faster AI services for customers and employees" — Chief IT Officer. Result: 99% accuracy achieved by customer-facing AI chatbot.

Focus

To ride on such momentum, SEA enterprises should focus on 5 strategic moves.

Five moves

5 strategic moves that chart the path to value.

Enterprises unlock 5x more AI value when strategic clarity meets applied AI.

Move 1 — Give AI leaders real authority, not just a title

• Appoint AI leaders at the enterprise and BU/ Function level with real mandate, accountability, and influence to drive value delivery
• Having AI leaders is not enough if the roles stay symbolic

Move 2 — Commit to bold topline ambitions, not just cost

• Set bold revenue targets – cost takeout is half the prize
• Cascade to BUs with clear execution KPIs
• Communicate targets and KPIs clearly with investors

Move 3 — Prioritize few, scaled plays over a graveyard of pilots

• Run AI workflows at scale with harnesses and guardrails in place
• Deploying AI doesn't automatically translate into value if deployment stays isolated

Move 4 — Build a real AI platform, not AI components

• Instead of betting on disparate components, build structural foundation that lasts: AI platform, data, and guardrails
• An agent-ready architecture is a "department store", not a single shelf

Move 5 — Supercharge employees to work in human-AI teams

• Don't tie employees' hands at work; equip them with the right tools, guidelines, and skills instead
• In Future-built teams, humans hand off toil to AI agents and focus on strategic steering

Move 01 — Leadership (part 1 of 2)

AI leaders need real mandate, accountability, and influence to drive value.

Maturity score on AI delivery model, from "siloed AI initiatives, team roles unclear" at the low end to "dedicated AI leadership teams with clear mandate" at the high end: Future-built average 96; SEA average 55.

Enterprises with Future-built AI delivery models realize 50% more revenue growth in 2025 vs. SEA average.

Quote: "Competitive advantage will require strong leader ownership…" — Digital HR Senior Director, Consumer Products.

Quote: "We must shift the mindset… Accountability needs to reside with the business" — Head of CEO Transformation Office, Banking.

Move 01 — Leadership (part 2 of 2)

The Future-built AI delivery model appoints AI leaders at both the enterprise and BU/ Function level to drive enterprise-wide clarity:

Enterprise C-Suite. Enterprise AI leader: shows real influence over the enterprise AI mandate; define enterprise-wide AI governance and guardrails; accountable for overall value outcomes.

BU/ Function 1 Leadership, BU/ Function 2 Leadership, BU/ Function 3 Leadership. BU/ Function AI leader: own AI deployment within BU/Function; accountable for BU-/Function-level value outcomes.

Move 02 — Targets

Future-built enterprises commit to bold topline ambitions, not just cost.

How SEA sizes the prize, on one scale: SEA enterprises put 25% of their revenue base and 30% of their operating cost base in scope for AI — sizing cost above revenue. Future-built enterprises put 42% of revenue and 41% of cost in scope — sizing revenue above cost.

What Future-built organizations do differently: 1. Set a bold enterprise-level AI ambition anchored in revenue outcomes, not just cost. 2. Translate enterprise ambition into clear execution KPIs for business units (e.g. time-to-market, adoption rate, project ROI). 3. Communicate the target clearly with investors and monitor continuously.

How a global bank led an ambition-driven transformation: 1. It set an ambition to be an AI leader targeting ~US$1b in profit and revenue uplift by 2027. 2. It cascaded that ambition into 10 "Big Rock" projects across BUs and supporting functions (e.g. wealth management, personal banking, IT), each needing to deliver US$100m+ in uplift. 3. It announced the ambition during its Investor Day in 2025.

Move 03 — Scale

SEA enterprises run some AI in production, but have yet to scale.

Maturity Score on AI Productization: SEA average 50, Future-built average 93, on a scale from “AI solutions remain prototypes or isolated deployments” to “AI products operate as managed, scalable services”.

Future-built enterprises are 7x more likely to have agentic AI live and paying off vs. SEA enterprises, and 2x more likely to have deployed AI workflows at scale vs. SEA enterprises.

The key is to prioritize few, scaled plays: cut pilots and reshape the highest value workflows, not just the simplest ones; scale AI beyond isolated sandboxes and business units, with harnesses and guardrails; accelerate adoption by sharing use cases, deployment lessons, and reusable playbooks.

“Whenever companies take a core workflow to scale (such as credit decisioning in banking and demand forecasting in consumer goods), it delivers double-digit gains in productivity, revenue, and cost reduction” ~ BCG The Formula for Agentic AI Value 2026.

Move 04 — Platform

Future-built enterprises build an agent-ready "department store" instead of betting on disparate components.

9 in 10 enterprises in SEA are not fully ready for AI agents, while Future-built enterprises have an agent-ready architecture:

The Agentic AI Department Store, from the top floor down: Floor 7, Engagement (One front door to AI, Ready-to-use agents, Search company data, Local languages); Floor 6, Agents (Agent templates, No-code builder, Developer framework, Agents working together); Floor 5, Models (Frontier, Workhorse, Efficient, Specialist: image, video, speech, code); Floor 4, Data + Tools (Company data, Enterprise search + RAG, App connectors, APIs + MCP tools); Floor 3, Build + Runtime (Managed agent hosting, Automatic scaling, Persistent memory, Fast set-up); Floor 2, Govern + Optimize (Agent identity, Policy checkpoint, Prompt guardrails, Observability + evaluation); Floor 1, AI Hypercomputer (ASIC, GPUs, Networking + storage, Serving, Flexible consumption).

Laying the groundwork means getting these right:

Agent-ready data layer. Start with core data capabilities. Data basics: Data structuring; Access controls & security; Metadata & standards. Agent stack: Agent tools & permissions; Retrieval-Augmented Generation (RAG) /grounding; Semantic layers.

Centralized AI platform. Integrate cloud-native AI tools with enterprise-owned layers. Example: How a Cross-Cloud Knowledge Catalog enables an Enterprise AI Platform: Custom Applications and Agentic Development on a Lakehouse, over the Knowledge Catalog (Universal Context Engine), with context federation to Cloud Storage and SaaS Applications.

• Unified context: connects multimodal enterprise-wide data into a single view
• Zero Copy Saas Ecosystem at Zero Cost: query live data without replication
• Enrichment: Automated, agentic enrichment with human-in-the-loop validation

Enterprise-wide guardrails. Implement key controls that drive agentic value. Statistical significance on agentic value, %: Memory & context handling 32%; Scope, autonomy, oversight 21%; Evaluation & rollback 18%; Tools, integration, API 13%; Cross-function ownership 9%; Security, audit 7%.

Move 05 — Human-AI teams

In Future-built teams, humans are supported by teams of AI agents to focus on strategic steering instead of managing fragmented systems.

Employees are already speed-charging their personal lives with AI: 78% are regular AI users; 50% save more than 1 hour a day by using AI.

…don't tie their hands at work; equip them instead. Essentials for human-AI teams: 1. Enterprise-grade AI tools and connectors. 2. Clear acceptable-use policies and no-fly zones for employee AI usage. 3. AI learning integrated into day-to-day operations, supplemented with structured training.

Quote: "Access to the right tools is just as important as the ways of using and governing them" — Chief IT Officer, TMT.

Today: more than 4 hours a week spent on toil-heavy tasks (BCG Making Work Work 2024) — frequent context switching across fragmented systems; manually sourcing, tracking, updating. Tomorrow: more than 53% of employees saving more than 1 day per week with reshaped workflows (BCG AI at Work 2026) — a team of specialized agents running specific tasks; a human in the loop steers and approves key decisions.

Employees shouldn't have to bring their own tools to work

At a glance

Southeast Asia's next leapfrog moment is here – sustained competitive advantage now lies where strategic clarity meets applied AI.

Future-built enterprises already realized 5x more value globally. 2025 value realized through AI, as a percentage of revenue, by degree of strategic clarity and applied AI maturity: low clarity and high maturity, 2.0% — mature capabilities drive results, but low clarity leaves value untapped. High clarity and high maturity, 4.5% — transformative impact where strategic clarity meets applied AI. Low clarity and low maturity, 0.9% — lack of clarity and capabilities results in very low impact. High clarity and low maturity, 1.8% — clarity on where value lies, but capabilities must evolve to realize it. Between the lowest and highest quadrants: 5x AI value. Source: BCG Applied AI Index 2026.

SEA enterprises' path to Future-built is now charted by 5 key moves, from an SEA average of 41 to a Future-built 76+: 1. AI leadership with real authority. 2. Bold revenue targets. 3. Prioritized scaled plays. 4. Centralized AI platform. 5. Human-AI teams.

Methodology

AI maturity is assessed across the value chain among participating enterprises in Southeast Asia (n=60+) and globally (n=1.4k+), across eight dimensions: Strategy, Innovation, Customer Experience, Operations, Technology, Data, Op Model, People.

Resulting in an overall maturity score along a 4-segment scale. 0–25 Stagnating: AI is absent or ad hoc, with no consistent strategy, ownership or scaled use cases. 26–50 Emerging: Foundational capabilities and early experimentation exist, but value stays local and unproven. 51–75 Scaling: AI is embedded in core workflows with governance and repeatable delivery, and value is measured. 76–100 Future-built: AI is a source of advantage, reshaping operations with an operating model built to compound it.

Disclaimer

This report was commissioned by Google.

This document has been prepared in good faith on the basis of information available at the date of publication without any independent verification. BCG does not guarantee or make any representation or warranty as to the accuracy, reliability, completeness, or currency of the information in this document nor its usefulness in achieving any purpose. Readers are responsible for assessing the relevance and accuracy of the content of this document. It is unreasonable for any party to rely on this document for any purpose and BCG will not be liable for any loss, damage, cost, or expense incurred or arising by reason of any person using or relying on information in this document. To the fullest extent permitted by law, BCG shall have no liability whatsoever to any party, and any person using this document hereby waives any rights and claims it may have at any time against BCG with regard to the document. Receipt and review of this document shall be deemed agreement with and consideration for the foregoing.

This document is based on a primary qualitative and quantitative research executed by BCG. BCG does not provide legal, accounting, or tax advice. Readers are responsible for obtaining independent advice concerning these matters. This advice may affect the guidance in the document. Further, BCG has made no undertaking to update the document after the date hereof, notwithstanding that such information may become outdated or inaccurate. BCG does not provide fairness opinions or valuations of market transactions, and this document should not be relied on or construed as such. Further, any evaluations, projected market information, and conclusions contained in this document are based upon standard valuation methodologies, are not definitive forecasts, and are not guaranteed by BCG. BCG has used data from various sources and assumptions provided to BCG from other sources. BCG has not independently verified the data and assumptions from these sources used in these analyses. Changes in the underlying data or operating assumptions will clearly impact the analyses and conclusions.

This document is not intended to make or influence any recommendation and should not be construed as such by the reader or any other entity. This document does not purport to represent the views of the companies mentioned in the document. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by BCG.

Acknowledgements

Mark Micallef, Southeast Asia Managing Director, Google Cloud. Kunal Talwar, Strategic Partnerships Lead, Google Cloud. Venugopal Desai, Head of Partner Marketing, Google Cloud. Anant Shivraj, Managing Director & Partner, BCG. Hanno Stegmann, Managing Director & Partner, BCG X. Prasanna Santhanam, Head of SEA Partnerships, BCG. Marc Roman Franke, Partner and Associate Director, BCG. Saransh Mehtani, Principal, BCG. Jing Lin Lee, Consultant, BCG.

Copyright © 2026 by Boston Consulting Group. All rights reserved.