ASEAN Embraces AI, But Hurdles Remain

Despite the initial lag in Artificial Intelligence (AI) implementations in the year 2021-2022, AI projects have risen by 80% in the year 2023 across all ASEAN countries. The budget is expected to increase by 67% for the year 2024-2025. While excitement around AI is booming in Southeast Asia, widespread and sustainable implementation faces significant challenges.

This article explores these roadblocks hindering large-scale AI adoption across various sectors in the region. The insights are based on discussions with multiple AI solution providers in Thailand predominantly.

Even with the rise in the number of projects, many projects remain confined to smaller scale. The most successful implementations involve data analytics, impacting areas like marketing, fraud detection, government-sponsored smart cities, etc. Notably, the financial sector has taken the lead in adopting data-driven AI solutions. However, many projects have not progressed beyond proof-of-concept or Minimum Viable Product (MVP) stages in other sectors.

Key Challenges for ASEAN Countries

There are several challenges in implementing AI in the ASEAN market. Some of these are discussed in the following sections.

• Scarcity of AI Expertise

  A significant barrier is the lack of readily available companies with large-scale AI capabilities within ASEAN. Existing providers are primarily located outside the region, with China being the closest hub. Similarly, consulting firms specializing in large-scale AI implementation are less, often lacking a strong focus on AI in their ASEAN branches.

• Talent GAP

  As with other regions, ASEAN faces a shortage of skilled personnel across the AI spectrum, from algorithm development to critical thinking about AI solutions. This talent gap is further amplified in specific countries like Thailand with older demographics, where reskilling challenges differ from younger, tech-savvy generations.

  Resources from consulting firms also have similar issues where the current skills are primarily geared toward traditional business.

• Geopolitical Consideration

  The dominance of US and Chinese companies in AI technology development raises geopolitical concerns. ASEAN nations must navigate these complexities to ensure ethical and sustainable AI implementation strategies.

  Any friction between the two countries has a direct impact on accessing AI technology or corresponding resources.

• Evolving Regulations

  AI-specific regulations are either absent or under discussion in many ASEAN countries. Current regulations rely heavily on the existing Personal Data Protection Act (PDPA) based on the European GDPR model. This localization process can be lengthy, taking over three years to implement, and may be further delayed in election cycles. Similar delays are likely once the EU finalizes its AI regulations. While Singapore boasts the most robust legal framework, harmonization across other countries is necessary.

  Aligning with Singapore's regulations could provide a degree of compatibility and safety in terms of implementing AI Use Cases.

• Investment Cost

  The expense of implementing AI solutions remains high in some ASEAN countries, mirroring trends in South Asia. Current cost structures often reflect US, UK, and Japan benchmarks. Without subsidies or belonging to a subsidiary of a major AI-implementing nation, achieving a positive return on investment (ROI) can be challenging. Additionally, local AI companies primarily cater to clients in developed regions due to higher margins, limiting their resources for domestic projects.

  The investment cost also needs to consider the changes that need to be done to Legacy applications to start getting the benefits of AI Solutions.

• Write-off Cost

  Many ASEAN retailers underwent significant digital transformations between 2017 and 2019. Large-scale Enterprise Resource Planning (ERP) implementations often involve substantial write-off costs, typically spread over a decade in some countries. The additional write-off burden associated with large-scale AI adoption discourages some companies, leading them to favor smaller-scale data analytics solutions. Justifying the ROI for AI projects remains a hurdle in ASEAN, further compounded by the region's lower resource costs compared to developed nations. This challenge extends to South Asian countries as well.

  This is very complex and there are many factors affecting the write-offs. However, one large retailer in Thailand had this number close to 20 million USD spread across eight years. The numbers are similar across large enterprises across sectors.

• Job Displacement

  AI could increase productivity by automating user tasks. But there are going to be significant disruptions to the workforce. McKinsey estimates that 23 million jobs could be displaced by automation by 2030. There are new jobs that will replace the old jobs. This depends on various factors like shifting labor demands, government regulations, cognitive skills, etc.

  There is no clear view of how the job market will shape up. Hence decisions around AI projects impacting the workforce might be slow.

Conclusion

Even though there are many AI projects, there needs to be a clear and flexible strategy to get the ROI out of the projects. The strategy needs to be flexible enough to adapt to different government regulations, which take time and are subject to change.

Ethical consideration is very important when handling the issue of job displacement in ASEAN countries.