Bridging the GenAI divide in ASEAN, Asia, and Singapore

The AI paradox: High adoption, low transformation

A gloomy picture of the enterprise AI landscape is provided by the "State of AI in Business 2025" study, which reveals a "GenAI Divide" in which 95% of organisations do not see a return on their large investments.  This conundrum is a reflection of a larger, worldwide investment trend rather than just a business issue. A self-reinforcing bubble that stifles other potentially revolutionary breakthroughs is being created by the growing concentration of VC and PE funding on AI businesses, especially those specialising in generative AI. This concentration is wasteful and unimaginative.

For the first time since 2015, financing for renewable energy start-ups decreased in 2023, despite AI attracting an astounding $50 billion in investment. Early in 2025, this pattern persisted, with AI start-ups obtaining a record 57% of worldwide venture capital funding, with a sizable amount going to a select few big deals. The report's findings unequivocally demonstrate that this infusion of capital into a limited number of AI solutions, frequently linked to a "fear of missing out" (FOMO), is causing overvaluation without a commensurate increase in P&L impact. The real value of AI is found in its capacity to enhance and deeply integrate into particular, problem-oriented workflows rather than in stand-alone, generic applications.

Navigating the GenAI divide in ASEAN and Singapore

The paper highlights that successful companies frequently "buy rather than build," collaborating with suppliers who provide tailored, learning-capable technologies. With a thriving AI start-up ecosystem, Singapore can establish itself as a strategic partner in the area. Rather than simply exporting general-purpose AI solutions, Singaporean companies can collaborate with companies in other ASEAN nations to create tightly integrated, process-specific AI tools. This is consistent with the report's conclusion that external partnerships outperform internal builds considerably (67% vs. 33%).

The analysis dispels the assumption that AI will instantly replace the majority of occupations, emphasising instead that its primary effect will be on external spending, including agency fees and BPO. This is a crucial realisation for ASEAN's emerging economies. Instead of replacing human labour, the goal should be to complement them using AI to automate monotonous, routine jobs and free up human creativity and inventiveness for higher-value, more complicated work. AI-powered precision farming, for instance, is expected to boost crop yields and save water consumption in the agricultural sector. AI is also being used by businesses like Gojek and Grab to power essential production services and optimise promotions.

In a digitally native region like Southeast Asia, the report's finding of a "shadow AI economy" in Western businesses, where workers use personal gadgets for work, strikes a deep chord. Organisations should "bridge this gap by learning from shadow usage and analysing which personal tools deliver value before procuring enterprise alternatives" because this unofficial adoption can be a potent source of insight. This bottom-up strategy is a tried-and-true method of ensuring operational fit and speeding adoption by enabling line managers and "prosumers" to identify issues and evaluate potential solutions.

Rethinking infrastructure and investment I: Beyond the hype

When considering global investment trends, the report's findings highlight serious concerns regarding the concentration of capital in data centres and artificial intelligence. The paper cautions against the "enterprise paradox," in which large companies lead in pilot volume but lag in scale-up. It also points out that, despite back-office operations having a greater return on investment, the majority of investment is now going into visible, top-line services like sales and marketing. Particularly in the building of data centres, this speculative trend is reminiscent of previous bubbles and poses a significant risk.

• The problem with latency and cheap energy: Findings from the analysis directly contradict the notion that data centres need to be situated near inexpensive energy sources. This "latency" argument, according to which the location of data centres is mostly determined by the cost of power, is getting out of date. A more distributed and sustainable "Agentic Web" is needed in place of the large, centralised data centre infrastructure architecture, which needs to be re-evaluated in light of AI's ascent. Off-grid sustainable power is becoming a strategic commercial necessity in addition to being an environmental necessity. In order to increase energy efficiency and make use of green energy sources like bioenergy and low-carbon hydrogen, Singapore has already started down this path with its Green Data Centre Technology Roadmap. 

If off-grid solutions are effectively implemented and fuelled by green or waste-derived energy (particularly from mining and agricultural wastes), data centres can overcome the dual problems of sustainability and latency. 

• An investment strategy for impact: The research emphasises that less obvious, back-office tasks like finance and procurement frequently offer the best return on investment. This means that investors should focus on technology that produce quantifiable, process-specific efficiencies rather than gaudy, front-end applications. Investing in less evident but very influential technologies is part of this. The transformation of coal tailings or agricultural waste into green fuels is a good example. In developing countries, this type of innovation has the potential to establish completely new, self-sustaining ecosystems that have substantial positive externalities for the environment in addition to a high internal rate of return.

• Contextualised solutions for the Global South: A lesson for working with the Global South can be learnt from the report's emphasis on deep customisation and a "build-to-buy" strategy for success. First World-developed generic, one-size-fits-all solutions frequently overlook local contexts, distinct procedures, and infrastructure constraints. Working together with regional partners to jointly create solutions that are suited to the unique requirements of a community or sector is a superior strategy. This is in line with the efforts of groups like the Global Sustainability Foundation Network (GSFN), which chairs blockchain projects and prioritises solutions that are contextualised..

Rethinking investment: A path beyond the AI bubble in ASEAN, Asia and Singapore

As mentioned previously, the erroneous belief that low latency and inexpensive energy are the main non-negotiables for AI is the driving force behind the current investment concentration on massive, power-hungry data centres. However, this paradigm is not only financially unstable but also unsustainable in terms of the environment. Investing in the core technologies that can enable a really robust and decentralised "Agentic Web" is a more progressive approach for ASEAN, spearheaded by Singapore.

By giving basic, sustainable technologies that are being overshadowed by AI hype priority, a deliberate change in funding can close the GenAI Divide and promote long-term balanced economic development (the term growth is often over emphasized). 

This includes:

• Photonic-electronic hybrid chips: These chips are essential for lowering energy usage and boosting speed at the edge since they convey data using light. Recently, a Chinese business collected more than 1.5 billion yuan ($200 million) for its photonic-electronic hybrid chips, which can lower power consumption by up to 30% and enhance calculation speed by 100 times. These new technologies are essential to the development of durable robotics and human-android surrogates. Significantly improved data centre capabilities with significantly reduced consumption characteristics will also come from further versions.

• Off-grid power and modularized waste-to-energy: Investing in modular waste-to-power generation can enhance the back-office ROI and lower external spend. This has been proven by Hong Kong's modular waste-to-energy plant, which produces enough electricity for 100,000 households while processing 1.1 million tonnes of waste yearly. Decentralised deployment of this technology throughout ASEAN could result in the development of off-grid, self-sustaining ecosystems with significant profits and favourable environmental externalities.

• Methane pyrolysis: Methane pyrolysis is a renewable energy method that emits no CO2 and yields solid carbon and hydrogen. Although there have been large investments in start-ups in this field, such as Tulum Energy, the total amount of capital still trails other clean hydrogen technologies. This is a high-impact, financially viable sector that can offer an affordable, scalable route to industrial decarbonisation. Singapore has also been deeply involved in the Low Carbon Energy Research (LCER) and investments into the pyrolysis ecosystem that has an equally useful off taker for its carbon fibre/nanotubes for graphene production is being explore as a complimentary carbon-circular ecosystem.

• Reforming hydrocarbon chains: In the Global South, technologies that can turn waste streams — like coal tailings and agricultural waste — into green fuels have the potential to establish whole new economic ecosystems. Although less evident to a software-focused venture capital landscape, these solutions provide a potent blend of social, economic, and environmental advantages. This will have a “turbo-charging” effect on economies as the reliance of fossil fuel import (foreign exchange balance) could be further improved, enabling better use of developmental funds without dipping into the often difficult global financial system that deems Global South financing a “greater risk”. 

Singapore as the catalyst for a sustainable AI Rethink

In ASEAN and beyond, Singapore is in a unique position to be the driving force behind a radical rethink of AI adoption. A deliberate, human-centred approach to technology and its established leadership in digital governance make it the perfect model for bridging the "GenAI Divide." Singapore has a strong foundation based on trust, legislation, and real-world application, in contrast to many countries who are racing into an AI arms race.

• Human-centric governance: Singapore sets a global norm for AI governance with its proactive approach. Singapore strikes a balance between innovation and accountability through programs like as the voluntary Model AI Governance Framework, which provides guidance to private sector organisations on ethical and governance problems. Instead than being a strict set of guidelines, this framework is intended to be a tool for improving corporate operations. Its voluntary character hasn't lessened its impact; major Asian organisations have adopted it as a norm, establishing de facto regional guidelines that frequently go above and beyond what is required by law elsewhere. Singapore “light touch” approach hopes to encourage innovation while stepping in to provide iteratively more robust governance to build trust and public assurance and not to hinder development – a fine line that Singapore has vast experience in navigating as a survival tool since our independence. 

• A "Living Lab" for AI solutions: Singapore's "Smart Nation" program offers a special environment for AI experimentation. When paired with regulatory sandboxes, this strategy enables the safe testing of cutting-edge AI technologies in a protected environment prior to widespread implementation. One practical example of how 5G and AI might spur change is the Tuas Port, which is utilising these technologies to create a fully automated logistics ecosystem.

• Bridging the regional gap: Although ASEAN as a whole falls behind in the adoption of AI and 5G, Singapore is a digital leader in the region. Through the ASEAN Digital Ministers Meeting (ADGMIN) and the Working Group on AI Governance (WG-AI), Singapore is spearheading initiatives and conversations aimed at closing this gap. Singapore helps the area avoid the traps of fragmented, uncoordinated growth by sharing its frameworks and best practices, which have an impact on the recent ASEAN Guide on AI Governance and Ethics. ASEAN member countries will too have to “trust the process” and not let antiquated thinking of the “big brother syndrome” from hampering progress via individual country’s comparative advantage – with Singapore being the obvious leader in digital implementation in ASEAN. 

• Focus on sustainability: The world's first sustainability standard for data centres in tropical regions was implemented in Singapore in recognition of the high energy usage of data centres. Singapore is positioned as a leader in the development of ethical and ecologically responsible AI because to its dedication to "Green AI" through collaborations, policies, and an emphasis on energy conservation. Recognising Singapore capabilities need not to see as “losing face” for other countries but as opportunity to exercise meritocracy in terms of multilateral support for the betterment of ASEAN. 

Singapore's strategy, which is distinguished by a principles-based, light-touch governance framework that promotes voluntary compliance and a cooperative environment, offers other countries a clear way to bridge the GenAI Divide. Singapore can take the lead in redefining AI for the world community by putting human well-being first, building trust, and making investments in sustainable practices.

Conclusion: A collaborative call to action

The "GenAI Divide" is a gap in methodology rather than technology. To overcome it, one must fundamentally change their perspective: from one that emphasises hype to one that is dedicated to real value; from one that is "buy-and-forget" to one that is "partner-and-co-evolve"; and from one that is narrowly focused on Western markets to one that is inclusive and collaborative for the Global South.

 
This implies the following for Singapore and the rest of ASEAN:

1. Prioritizing sustainable infrastructure: Invest in distributed, sustainable energy solutions to lessen your influence on the environment and create a more robust "Agentic Web" instead of depending solely on cheap, centralised electricity. With its Green Data Centre Technology Roadmap, Singapore has already started down this path.

2. Embracing complementary AI: Understand that the real benefit of AI is that it may enhance human creativity rather than replace it. Pay attention to solutions that increase back-office productivity and empower local workers.

3. Building trust through collaboration: Give up using a one-size-fits-all strategy. Collaborate with regional partners to jointly develop trust-based solutions that are intricately woven into particular workflows. The ability of Singapore to serve as a regional bridge for talent, technology, and sustainable practices will determine its viability as a centre.

4. Leveraging decentralized innovation: Seek information from the "shadow AI" economy and enable frontline managers to advocate for change. The most significant changes will probably be made from the bottom up rather than the top down.

A unified ASEAN strategy

A fresh, cohesive approach is needed to realise this ambition and steer clear of the dangers of a speculative AI bubble. For ASEAN, we suggest a cooperative structure that:

• Aligns public and private capital: Allocates some of the enormous sums of money going into AI to sustainable, basic solutions that tackle the particular problems facing the area.

• Fosters cross-border partnerships: Promotes joint development and implementation of tailored, learning-capable AI solutions between Singaporean IT firms and companies in other ASEAN countries. This endeavour can foster closer inter-ASEAN consensus and the common-ASEAN-mindset. 

• Champions contextualized solutions: Emphasises spending on dispersed and modular technologies that can build highly effective, self-sustaining ecosystems in the Global South, like waste-to-energy and off-grid power. For improvement and comprehension of the many needs of ASEAN communities, this is also a fantastic feedback mechanism.

By adopting this strategic change, ASEAN can set the global standard for creating a robust, just, and sustainable AI ecosystem, paving a path that will benefit the economy and the environment greatly.

The window of opportunity to bridge the GenAI Divide is closing quickly. According to the report, companies who do it correctly will create long-lasting moats built on memory, learning, and deep integration, as well as lock in vendor relationships. Instead of copying a bad Western paradigm, AI in Asia should steer towards a new path that is both profitable and incredibly beneficial for both people and the environment.

illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy writers, their opinions do not necessarily represent those of illuminem.

Track the real‑world impact behind the sustainability headlines. illuminem's Data Hub™ offers transparent performance data and climate targets of companies driving the transition.