Today in Bangkok, 50 disaster management leaders from South and Southeast Asia met to turn interest in AI into tools you can actually use during emergencies.
How can you speed up critical decisions when every minute counts? That was the question guiding the AI Jam for disaster response professionals, organized by OpenAI with the Gates Foundation, ADPC and DataKind.
What happened in Bangkok
Representatives from 13 countries participated: Bangladesh, India, Indonesia, Lao PDR, Malaysia, Myanmar, Nepal, Pakistan, the Philippines, Sri Lanka, Thailand, Timor-Leste and Vietnam.
They came from government agencies, multilateral bodies and NGOs on the front lines coordinating information and supporting affected communities. Think of the people who answer calls, write situation reports, and decide where to send the next truck — that was the kind of experience in the room.
During the day, teams worked side by side with mentors from OpenAI to find practical AI uses that can fit into real processes. Instead of talking about theory, they explored building custom GPTs and reusable workflows for concrete tasks: situation reports, needs assessments, public communication and other operational activities.
Why this matters now
Asia is the world’s most disaster-exposed region. Roughly 75% of people affected by catastrophes globally live in Asia. The World Bank estimates disasters have cost ASEAN countries more than $11 billion in recent years.
There are also signs people already turn to AI in crises. During Cyclone Ditwah in Sri Lanka, related messages on ChatGPT jumped 17-fold. In Thailand, during Cyclone Senyar in November 2025, message volume rose 3.2 times compared with previous months.
Isn't that a clear sign that AI can play a direct role in how people seek help and get information? If you’re wondering whether it’s relevant, those spikes should get your attention.
What the AI Jam tested and what they learned
- Practicality was prioritized: build models and scripts teams can use with limited infrastructure.
- There was a focus on institutional trust and responsible use: it’s not about automating everything; it’s about integrating AI where it adds value without undermining critical processes.
- They worked on modular solutions: custom GPTs to generate situation reports, standardize questions for needs assessments, and draft clear public messages.
Participants especially valued being able to iterate on prototypes during the session instead of receiving theoretical materials. That speeds up adoption because teams leave with something they can try in the field — a working draft, not just a slide deck.
This meeting aims to close the gap between what AI can do and its real use in the field. The opportunity is in turning interest into practical capacity. - Sandy Kunvatanagarn, OpenAI
Equipping those closest to communities with digital skills is a key investment in preparedness and response. - Dr. Valerie Nkamgang Bemo, Gates Foundation
Combining AI tools with regional expertise can strengthen early warning systems and risk analysis. - Aslam Perwaiz, ADPC
Where they're headed next
OpenAI and partners are evaluating a second phase focused on deployment pilots and deeper technical collaboration. The idea is to support these organizations as they implement and adapt tools to local conditions.
This isn’t a magic solution. But if AI is designed with real contexts in mind — fragmented data and limited resources — it can speed up information gathering, improve coordination and help you make faster, better-informed decisions.
If you work in disaster response, think of AI as a tool to amplify what you already do well: communicate clearly, prioritize needs and respond with speed.
Original source
https://openai.com/index/helping-disaster-response-teams-asia