AI Agents, Uneven Futures: Literacy, Risk, and Governance in Asia-Pacific

NetMission Ambassador Blog Posts, NetMission Digest

Welcome to Issue #7 of NetMission Digest. As we move further into 2026, the digital landscape is undergoing a fundamental shift. While the previous years were defined by generative AI, Microsoft leadership labelled 2026 the year of  AI agents – systems that, unlike chatbots that respond to prompts, can independently plan, make decisions, and execute tasks; booking, coding, negotiating, with minimal human input. This shift raises urgent Internet Governance questions: who sets the rules for these systems, who is held accountable when they fail, and which countries get to shape these standards? 

However, as these agents become embedded in our economies, a critical Internet Governance challenge is emerging: the enthusiasm for agentic AI is spreading unevenly, creating new layers of risk and digital divide across the Asia-Pacific region.

The “Next Great Divergence” in Asia-Pacific

A recent flagship report from the United Nations Development Program (UNDP) warns that we are entering an era of the “Next Great Divergence.” While trade and technology previously helped close global gaps, the uneven adoption of AI now threatens to widen inequality between nations. The statistics in the Asia-Pacific region tell a story of two realities: in some high-income economies, nearly 2 in 3 people already use AI tools, whereas in many low-income states, usage remains closer to 5%. This gap is not just about access to tools; it is about uneven readiness in terms of skills, connectivity, and governing systems. Countries with strong regulation and compute are poised to capture a massive “AI dividend,” while others face higher vulnerability to job disruption and data exclusion. This is not only a development gap. It is a governance power gap. Countries that cannot build or regulate AI systems become rule-takers rather than rule-makers, with limited voice in the global standards that will shape their digital futures.

From Innovation-First to Sovereignty-First: Contrasting Governance Approaches in APAC

Japan serves as a key entry point for understanding this trend. Under its  “Society 5.0” vision, Japan is leveraging AI to address complex social challenges like aging demographics. The Japanese government’s Artificial Intelligence Basic Plan (2025) explicitly aims to make Japan “the most AI-friendly country in the world” to avert a projected ‘2025 digital cliff,’ a scenario where slow AI adoption triggers significant economic loss through reduced productivity and competitiveness.

A core part of this strategy is the rise of AI agents that can autonomously manage workflows, robotics, and even conduct transactions with each other in an emerging “AI economic sphere.” By focusing on a “light touch” regulatory approach, Japan hopes to foster an environment where AI innovation flourishes through voluntary risk mitigation by businesses. This innovation-first, light-touch approach contrasts with efforts elsewhere in the region, particularly among ASEAN nations, that prioritize sovereignty and cultural relevance over speed, as explored in the literacy section below.

New Risks: Vibe Coding and Shadow AI

This rapid shift introduces governance challenges that share a common root: regulation is lagging behind deployment, leaving accountability gaps that youth advocates must monitor closely.

  • Shadow AI and Security: The unmanaged use of AI agents, often referred to as “Shadow AI,” can lead to unverified systems handling sensitive data. Autonomous agents can also be used by malicious actors to conduct more sophisticated cyber-attacks or generate vulnerable software code 
  • Vibe Coding and Literacy: We are seeing the rise of “vibe coding.” It means using natural language instead of computer code to build applications. While this democratizes access, it creates an Internet Governance issue: who is responsible when a natural-language prompt generates insecure or biased code?
  • Labor and Equity: The UNDP notes that female employment is nearly twice as exposed to AI displacement as male employment in the region.
  • The Reality Crisis: Deepfake misuse has surged, with a 1,500% increase in cases in APAC between 2022 and 2023, eroding basic trust in digital content, as explored in our previous issue.

Regional Responses: Literacy as Governance Infrastructure

Governments across Asia-Pacific are responding with targeted literacy and safety programs – recognising that AI literacy is not just an education issue, but a prerequisite for meaningful participation in AI governance. Without it, communities cannot hold systems accountable or influence the rules that govern them:

Conclusion

The rise of AI agents is more than a tech trend; it is a question of who governs the systems shaping our economies, our labour, and our digital lives. As youth leaders in the Asia-Pacific, we are not just end-users — we can be advocates in the multistakeholder forums, policy consultations, and regional bodies where these rules are being written. Moving beyond ‘how to use’ tools to ‘how to govern’ them is how we ensure that the ‘Next Great Divergence’ does not harden into a permanent divide.

Further Reading

NetMission Digest – 2026: Issue #6 (May 11, 2026)
Written by Yukako Ban (Reviewed by Jenie Fernando and Sherry Shek)