Asia Pacific’s startup ecosystem is highly dynamic, with continuous growth and evolution. But the need for equitable access is a critical issue for the region’s future, especially with AI at the forefront of innovation.
Startups in the region face significant AI infrastructure, talent and funding challenges. In fact, just in Southeast Asia alone, the anticipated growth still requires a massive investment of US$60 billion in infrastructure over the next 5 years.
What can be done to clear the hurdles, and what exciting developments may be in store for AI start-ups in Asia Pacific? DigiconAsia finds out more from Mona Li, Global Head of Startup Ecosystem, Nebius.
What are the key challenges and opportunities for start-ups in the region?
Mona Li (ML): As I work closely with early-stage companies across Asia Pacific, it’s clear that startups here face a unique duality – immense potential paired with very real constraints.
The infrastructure gap is perhaps the most urgent challenge with more than half of enterprises across the region still lacking the digital backbone needed to support AI development. Unlike tech giants, startups don’t have the luxury of building out in-house AI infrastructure, and as funding becomes more selective, access to high-intensity computing infrastructure becomes even harder.
We see this acutely in Southeast Asia, where venture capital funding has declined. It’s not just about capital; it’s about ensuring that capital translates into meaningful access to the tools that drive AI innovation.
Talent is another persistent hurdle. The demand for AI engineers and technical talent continues to outpace supply, especially in emerging markets. At the same time, regulatory complexities across different jurisdictions, ranging from data sovereignty to AI ethics adds another layer of friction for founders trying to scale across borders.
And beyond the urban centers like Singapore, many promising teams simply don’t have equitable access to the digital resources required to build and compete globally.
That said, this is also a moment of extraordinary opportunity. Cloud-native AI infrastructure is rapidly leveling the playing field. At Nebius, we’re working to bring down the cost and complexity of accessing GPU clusters so that startups can focus on building and scaling, not budgeting. Open-source AI models are accelerating this shift with tools like DeepSeek V3 allowing startups to fine-tune advanced models for local use cases without breaking the bank.
Governments are already stepping in with smart, targeted policies. Singapore’s Enterprise Compute Initiative is a standout example of how public-private alignment can lower systemic barriers.
What excites me most is the sector-specific innovation we’re seeing, from healthtech and fintech to edtech and gaming. These are areas where startups can build fast, solve real problems, and scale across borders. Not to mention, APAC’s diversity isn’t a challenge, but a competitive advantage
As more founders build solutions that reflect the complexity of the region, emergence of globally relevant, deeply localized AI products will surface. Singapore may be the launchpad, but the impact will ripple across the region.
What are the biggest predictions for APAC’s AI ecosystem in the next 3-5 years?
ML: The next three to five years for AI in Asia Pacific will be defined by extraordinary transformation, only if the fundamentals are addressed. The biggest shift will come from startups moving away from building general-purpose tools to developing vertical AI solutions tailored for specific industries.
Whether it’s diagnostic intelligence in healthcare or hyper-personalized financial services, the companies that succeed will be the ones solving real-world problems with deeply contextual AI.
Affordable and accessible AI infrastructure will be the backbone of this transformation. Innovation will be spurred on by builders across different geographies who are not bogged down by rigid pricing models or closed ecosystems.
AI will become a core growth engine for national digital economies. In markets like Vietnam, Indonesia, and the Philippines, AI won’t just be a tech upgrade, it will be a way to leapfrog infrastructure gaps and deliver real economic outcomes. Startups will play a central role in this, provided they have access to the compute, capital and talent they need.
Another area I’m watching closely is the evolution of agentic AI. We’re moving past just generating outputs to building systems that can reason, adapt, and take action autonomously. This will reshape how software is built, and it opens up new opportunities for developers to create tools that are more dynamic, more responsive, and more cost-effective to scale.
Finally, we’ll witness decentralization of innovation. While Singapore remains a powerful hub, emerging ecosystems in other parts of the region are gaining momentum. With access to scalable cloud tools and open-source models, innovation is no longer confined to capital cities. It’s an exciting time for the region, and if we get the infrastructure right, the next wave of global AI leaders could very well come from here.
