OpenAI Lands in Bangkok: 50 Disaster Leaders Build AI Tools That Could Save Lives

OpenAI Lands in Bangkok: 50 Disaster Leaders from 13 Asian Nations Build AI Tools That Could Save Lives

When Cyclone Mocha tore through Myanmar in 2023, aid workers scrambled to piece together field reports arriving in six languages, three file formats, and zero standardised templates. Three years later, OpenAI gathered 50 of those same frontline responders in a Bangkok hotel ballroom and handed them something new: the tools to make sure next time is different.

The company's inaugural AI Jam for Disaster Management, held on 29 March 2026 in partnership with the Gates Foundation, the Asian Disaster Preparedness Center (ADPC), and DataKind, marks the first time a major Western AI lab has staged a hands-on capacity-building workshop specifically for Asia's disaster response community. It is a quiet but significant step in a region where 75% of all people affected by natural disasters worldwide live.

Why Bangkok, Why Now

Asia is not just disaster-prone; it is uniquely vulnerable to the cascading failures that follow a crisis. Fragmented communications infrastructure, multilingual populations, and under-resourced government agencies mean that the gap between data collection and decision-making can stretch from hours to days, precisely when speed matters most.

ADPC, which has coordinated regional preparedness since 1986, identified AI-assisted situation reporting as a top priority after reviewing response times across ASEAN member states. During a cyclone that struck Thailand last year, AI-related search queries from disaster responders surged threefold. When a separate cyclone hit Sri Lanka, that figure spiked 17 times.

The question we started with was simple: can AI help disaster response institutions react faster and more effectively when lives are at stake? The answer, after a day with these 50 leaders, is unambiguously yes.

— OpenAI spokesperson, AI Jam closing remarks

What the Participants Built

Representatives from Bangladesh, India, Indonesia, Lao PDR, Malaysia, Myanmar, Nepal, Pakistan, the Philippines, Sri Lanka, Thailand, Timor-Leste, and Vietnam spent the day working alongside OpenAI mentors. The outcomes were deliberately practical, not academic.

Teams built custom GPTs designed to:

• Summarise fragmented field messages arriving in multiple languages into unified situation reports
• Convert satellite and earth observation data into plain-language risk assessments
• Generate needs-assessment templates pre-populated with real-time weather and population data
• Automate public communication drafts calibrated to local languages and cultural context
• Improve early warning system accuracy by cross-referencing historical disaster patterns

We are not building toys. These are reusable workflows that a provincial emergency officer in Lao PDR can deploy the next time a flood warning hits. That's the bar.

— DataKind representative, AI Jam workshop

A Region That Cannot Afford to Wait

ASEAN alone absorbs over $11 billion in disaster-related losses annually, a figure that climate models suggest will rise sharply over the next decade. The AI Jam's focus on institutional trust, not just technical capability, reflects a growing understanding that adoption in government agencies requires more than a flashy demo.

By The Numbers - 75%: Share of global disaster-affected people who live in Asia (UNDRR) - $11 billion: Annual disaster-related economic losses across ASEAN (ADPC, 2025) - 13: Countries represented at the Bangkok AI Jam - 50: Disaster response leaders trained in a single day - 17x: Spike in AI-related queries from Sri Lanka during its 2025 cyclone

The Trust Equation

OpenAI's decision to partner with ADPC and the Gates Foundation rather than launching unilaterally speaks to a shift in how Silicon Valley AI labs approach emerging markets. The company is not the only Western firm betting on Asia's AI appetite, but this initiative targets public-sector resilience rather than enterprise revenue.

The workshop emphasised responsible use, data sovereignty, and building institutional confidence, themes that echo China's own approach to AI governance and the broader ASEAN conversation about AI readiness.

OpenAI confirmed it is exploring a second phase in the coming months, focused on pilot deployments and deeper technical collaboration with participating organisations. If successful, the Bangkok model could be replicated across the region, from Singapore's dealmaking corridors to the classrooms of South Korea.

What Comes Next for AI and Disaster Preparedness

The AI Jam is not a silver bullet. Participants acknowledged that connectivity gaps, budget constraints, and political will remain formidable barriers. But the workshop produced something tangible: a library of reusable GPT-based workflows that any ASEAN disaster agency can adapt without needing a dedicated AI team.

For a region where the next major disaster is a statistical certainty, that library could prove more valuable than any keynote speech.