Why ASEAN's diversity could be its biggest AI advantage

Southeast Asia's diversity, with multiple languages, cultures and legal systems packed into one region, is usually framed as an obstacle to harmonising artificial intelligence (AI) governance.

But the former United Nations (UN) practitioner, Dr Cornelia C. Walther, who is currently the associate professor at Sunway University in Malaysia, believes ASEAN's plurality could be turned into a strategic advantage.

“The cultural and linguistic diversity in ASEAN is a huge asset because it allows us to make a new type of AI that is far more versatile and far more inclusive,” she tells GovInsider.

Designing AI governance for plural societies pushes governments and institutions to move beyond one-size-fits-all solutions.  

Instead of importing systems designed elsewhere, ASEAN has a chance to shape AI that reflects the communities it actually serves, she says.

As she highlights, “how can we harness and bring out the best in our people and in our eco-environment?”

This forms the foundation of what Walther advocates as “ProSocial AI”.

Leveraging ASEAN's diversity for human-centred innovation 

ProSocial AI envisions a future that is pro-people, pro-planet and pro-potential.

Such a future recognises that technological progress and societal well-being are not competing priorities, but mutually reinforcing ones.

Many discussions around AI governance focus on preventing harm through transparency, accountability, and safeguards.

ProSocial AI starts with a different question: how can AI be designed from the beginning to actively improve people's lives?

"ProSocial AI is deliberately tailored, trained, tested and targeted to bring out the best in and for people and the planet," Walther explained.

ASEAN's current landscape reflects this tension as member states progress with different approaches to AI governance.

Singapore has adopted an adaptive approach through its Model AI Governance Framework; Malaysia has introduced the AI Governance and Ethics framework; Vietnam has moved towards dedicated AI legislation, while some other member states are still developing their respective frameworks.

“It's not enough to simply mention these principles in legislation,” Walther said. “Governments need to ask what they actually mean in practice and how to deliberately build them into AI systems.”

Redefining success in the public sector through ProSocial AI

Public sector AI initiatives are often measured through productivity gains, operational efficiency, and cost savings.

Yet these metrics may not fully capture whether the AI adoption truly serves the interest of civil servants.

Additionally, ASEAN governments are adopting AI at different levels of digital maturity.

The region’s challenge is not the speed of deployment, but ensuring that AI systems continue to benefit its diverse communities.

This is where ProSocial AI bridges the gap. Rather than evaluating AI solely by technical performance, the framework examines whether systems are designed with human and planetary outcomes in mind.

It is built around 4Ts principles:

• Tailored: AI solutions must address specific societal challenges and adapted to local contexts, languages, cultural norms as well as constraints.
• Trained: The dataset should be represented by diverse voices and languages to ensure the AI function equitably across different contexts.
• Tested: Rigourous testing should be done to examine potential biases and edge cases that might affect vulnerable populations
• Targeted: Ultimately, the solution focuses on measurable outcomes such as reducing carbon emissions and improving the collective well-being

To translate these principles into practice, Walther developed the ProSocial AI Index, which allows organisations to assess AI systems before deployment.

As ASEAN governments expand their AI investments, public institutions must therefore consider whether these systems serves the humanity, especially when innovations tend to overlook broader societal trade-offs like environmental impacts and institutional resilience.

"How can we move beyond the traditional, very narrow scope of return on investment towards a more holistic understanding of return on values?" she asked.

Walther noted that the ProSocial AI Index is currently being adopted by the Moroccan government, with pilots underway at both state and departmental levels.

Last December (as per a Google-translated from French article by United Nations Population Fund Morocco), the government had several exchanges with international experts and development actors to explore possible AI applications in the areas of social inclusion, reproductive health, youth and women's empowerment, vulnerability reduction, and environmental protection.

'Double literacy' needed for effective AI governance

As governments invest in AI literacy programmes, Walther argues that teaching people how to use AI is only half the equation, since effective AI governance cannot rely solely on technical expertise or regulatory compliance.

"We need to understand our artificial assets, but we also need a more holistic understanding of ourselves," she says.

She calls this double literacy: combining algorithmic literacy with human literacy.

Human decision-making is shaped by cognitive biases, emotions, social contexts, and environmental factors.

As AI systems become increasingly sophisticated, understanding human behaviour becomes just as important as understanding the technology itself.

Hence, civil servants should develop the capacity to critically evaluate AI outputs, understand their limitations, and exercise independent judgement.

“You can only be fully accountable for something that you have understood,” Walther adds.

Bridging education and public institutions

Double literacy should be embedded across schools, universities, and workplaces as quickly as possible.

Rather than waiting for legislation or large-scale curriculum reforms, Walther believes public institutions, universities and private organisations should begin equipping individuals with the skills needed to thrive in a hybrid human-AI future.

“Future generations will grow up with AI readily available at their fingertips, where cognitive tasks can be delegated almost instantaneously.

“While this presents unprecedented opportunities, it also raises important questions about critical thinking, creativity and human agency.” she notes.

For civil servants, the shift towards ProSocial AI does not begin with building larger AI models or drafting longer governance frameworks. It begins with the everyday decisions surrounding procurement, implementation and evaluation.

When selecting an AI system, organisations could ask four simple questions:

• Purpose: Does it solve a meaningful public problem?
• People: Does it strengthen rather than replace human judgement?
• Profit: Does it create long-term value for society?
• Planet: Does it minimise environmental and social costs?

As ASEAN accelerates the ambition, Walther believes the region has an opportunity to chart a different course.