Artificial intelligence is no longer a distant concept for Southeast Asia. From transforming public services to reshaping small businesses, AI, particularly generative AI, is steadily gaining ground across the region. But with this progress comes a pressing concern: will AI drive inclusive growth or will it deepen the existing gaps between digital haves and have-nots?

The region’s development has never followed a single trajectory. Major cities are racing ahead with AI integration, while rural and underserved areas are still battling fundamental challenges in digital infrastructure and literacy. 

We explore closing the digital divide in rural Southeast Asia with tech-powered initiatives and solutions

If AI is to truly uplift Southeast Asia, its adoption must be intentional, inclusive and rooted in the realities of each community.

A region of contrast: Digital haves and have-nots

Interestingly, Southeast Asia’s uneven digital landscape presents both opportunities and challenges. For example, countries like Malaysia, Singapore and Thailand are actively integrating AI into public service delivery and enterprise systems. But, on the flip side, rural communities in the Philippines, Myanmar and Indonesia are still grappling with limited digital literacy and basic internet access. To put that into perspective, let’s take Malaysia, half of its 34 million population living outside the country’s industrial and urban centres are making do with poor internet connectivity.

While the disparity is on the bandwidth or devices, it is also about the capability. In many parts of the region, digital literacy is low, so AI concepts, which are still relatively new, still remain abstract. If this whole AI wave is to benefit Southeast Asia equally, just like the rest of other parts of the world, it must address the foundational gaps in education and access.

Will localisation be the region’s biggest equaliser?

One of the most promising ways to bridge the divide is through localisation. This goes beyond just translating interfaces into Bahasa Melayu or Tagalog. It means planning and creating AI models that understand local dialects, social contexts, as well as cultural nuances.

Startups across the region, such as Mesolitica, launched the first Bahasa Melayu generative AI model on the Amazon platform, and others are also experimenting with localised language models that understand region-specific slang and idioms and can process mixed-language prompts.

Startups as bridges, not just builders

While governments do set the rules and policy frameworks, we often see startups that are leading the charge on AI democratisation today. For example, in the Philippines, they are leveraging generative AI to build chatbot teachers for underserved schools with limited staff.

So, these tools are designed to work in low-bandwidth environments and are trained on local curricula. On top of that, this is also a reminder that AI innovation doesn’t have to mean cloud-scale rollouts; it can also mean low-tech solutions designed with the inclusion of local constraints.

Another thing that we can give startups a pat on the back for is that they are also playing a huge role in building AI literacy. AI-powered tools like ChatGPT are allowing people to help with their daily operations. These can be the basic initiatives, and such grassroots efforts are proving more effective in driving adoption than top-down policies alone.

In order to launch more complex tools, basic tools must be accessible for people to try and error for themselves, so it gives them the opportunity to explore them first, rather than bombard them with such complex tools.

Can big tech players level the playing field?

The short answer is yes. But we need to understand that commitment from these big tech firms to be inclusive is a major key point. For instance, major players such as NVIDIA and Google are investing heavily in AI cloud infrastructure across Asia. Sure, these investments will unlock capacity for the region, but they must also include public-private partnerships that prioritise access and affordability in order to move forward.

Southeast Asia stands at a crossroads on this issue as AI may become the great equaliser such as improving access to healthcare, bringing quality education to rural communities and powering SMEs. At the same time, we could also argue that AI could also deepen existing divides that the region is already battling with, increasing growth for the Tier 1 cities, while sidelining other parts of the region.

What is clear is that AI adoption cannot follow the same playbook and have one set plan, as broadband or mobile rollout. When we talk about growing in the region, it does require nuance, localisation and a community-first mindset, which a lot of these big tech players fail or lack understanding of.

So, we do see that the digital transformation in the region often happens in leaps, not gradual shifts and that leap must include everyone. Both the challenges and opportunities are real. And the main question is no longer whether AI can change Southeast Asia, but whether it will change it for everyone.