Helping disaster response teams turn AI into action across A…

今天在曼谷，我们将来自东南亚和南亚的 50 位灾害管理领袖召集在一起，举行首届 AI Jam for Disaster Management ，该活动由 Gates Foundation 、 Asian Disaster Preparedness Center (APDC) 和 DataKind 联合举办。

这一行动直指一个既简单又紧迫的问题：在关键时刻，如何利用 AI 帮助政府和非营利组织更快、更有效地应对灾害？

与会者来自 13 个国家——孟加拉国、印度、印度尼西亚、老挝人民民主共和国、马来西亚、缅甸、尼泊尔、巴基斯坦、菲律宾、斯里兰卡、泰国、东帝汶和越南，代表政府机构、多边组织和非营利组织。很多人直接参与一线灾害响应，负责信息协调、支持受灾社区并在时间紧迫时做出决策。

这次活动也基于我们在达沃斯宣布的 OpenAI for Countries Program 的推广。其核心目标是帮助各类组织把对 AI 的兴趣转化为可在现实运营中使用的能力，把 AI 嵌入到它们每天面临的实际挑战中。

应对亚洲日益增长的灾害风险

灾害响应队伍经常在资源匮乏的环境中工作，面对碎片化的数据、繁琐的手工流程和有限的基础设施。这些限制会拖慢协调速度并延误关键决策，尤其是在信息时效性至关重要的快速变化情境中。越来越多的团队正在探索 AI 如何更好地支持这些工作流。

这种紧迫感正在加剧。去年下半年，一系列台风和强风暴袭击南亚和东南亚，扰乱了社区生活，并把灾害响应体系推到极限。亚洲仍是全球受灾最严重的地区，估计占全球受灾人数的 75%。世界银行估计，过去几年灾害已使 ASEAN 国家损失超过 110 亿美元。

与此同时，人们寻求援助的方式也在变化。在斯里兰卡的 Cyclone Ditwah 期间，内部数据显示与台风相关的 ChatGPT 消息量增长了 17 倍，表明在危机时刻人们已经开始利用 AI 获取信息与指导。2025 年 11 月的 Cyclone Senyar 中，泰国也出现了类似的 AI 使用激增，消息量较此前几个月增长了 3.2 倍。这些迹象显示，将 AI 更直接地融合到响应团队的信息收集、决策和沟通中，有着明显的机会。

构建务实的 AI 解决方案

这正是本次 Jam 的重点。今天的活动中，与会者与 OpenAI 导师并肩工作，寻找 AI 在日常工作中可实现的支持方式。他们并非从零开始，而是探索如何构建定制的 GPT 和可复用的工作流程，应用于不同场景——从态势报告到需求评估再到对外沟通。会议同时强调负责任使用 AI 以及在机构内部建立信任的重要性。

“本次会议旨在填补 AI 能力与现场实际应用之间的鸿沟。整个亚洲对 AI 的热情很高，但真正的机会在于把热情变成实用能力。通过与灾害响应专业人员直接合作，我们可以确保这些工具既有用、可获得，又建立在现实需求之上。”—— Sandy Kunvatanagarn ， Head of Public Policy at OpenAI 。

“把数字工具和像 AI 这样的新兴技术的知识与技能带给最靠近社区的人，是我们在灾害预防与响应方面能做出的最有力投资之一。我们很自豪能把区域内的合作伙伴汇聚在一起，并将这些努力转化为可以立即投入使用的工具。”—— Dr. Valerie Nkamgang Bemo ， Deputy Director, Emergency Response at Gates Foundation 。

“AI 为我们理解和应对灾害开辟了新可能。ADPC 将 AI 融入地理空间工具和风险分析，把卫星与地球观测数据转化为可操作的洞见。AI Skills Jam 有助于提升 AI 素养，赋能人们找到应对灾害的解决方案。

我们可以把 AI 工具与区域专长和伙伴关系结合起来，强化预警系统、改进风险绘图，并支持各国与社区做出更快、更有信息支持的决策。”—— Mr. Aslam Perwaiz ， ADPC Executive Director 。

接下来，我们与合作伙伴将探讨第二阶段计划，重点在未来几个月进行试点部署并与区域内参与组织展开更深度的技术合作。我们期待继续推进这项工作，打造切实可用的工具，帮助社区更好地备灾与应对灾害。

Today in Bangkok, we’re bringing together 50 disaster management leaders from across Southeast and South Asia for our inaugural AI Jam for Disaster Management professionals, in partnership with the Gates Foundation, Asian Disaster Preparedness Center (APDC) and DataKind. 

The question guiding this initiative is simple, but urgent: How can AI help governments and nonprofits respond faster and more effectively when it matters most? 

Participants come from 13 countries—Bangladesh, India, Indonesia, Lao PDR, Malaysia, Myanmar, Nepal, Pakistan, Philippines, Sri Lanka, Thailand, Timor Leste, Vietnam—representing government agencies, multilateral organizations and non-profits. Many are directly involved in disaster response on the ground, coordinating information, supporting affected communities, and making time-critical decisions.

This effort also builds on the expansion of our OpenAI for Countries Program⁠ announced at Davos. At its core, this work is about helping organizations move beyond interest in AI and into real-world applications, embedding it into the operational challenges they face every day. 

Responding to growing disaster risks in Asia 

Disaster response teams often operate in resource-constrained environments, working with fragmented data, manual processes and limited infrastructure. These constraints can slow coordination and delay critical decisions, especially in fast-moving situations where timely information is essential. Many teams are now exploring how AI can better support these workflows.

That urgency is only growing. In the second half of last year, a series of typhoons and severe storms across South and Southeast Asia disrupted communities and stretched disaster response systems to their limits. Asia remains the world’s most disaster-prone region, accounting for an estimated 75% of people affected by disasters globally. The World Bank estimates disasters have cost ASEAN countries more than $11 billion in previous years. 

In these moments, we’re also seeing a shift in how people seek support. During Cyclone Ditwah in Sri Lanka, internal data showed a 17× increase in cyclone-related messages on ChatGPT, highlighting how AI is already being used to access information and guidance during crises. During Cyclone Senyar in November 2025, Thailand saw similar AI usage surges, with message volume jumping 3.2× compared to the months prior. This points to a clear opportunity to integrate AI more directly into how response teams gather information, make decisions, and communicate during emergencies.

Building practical AI solutions

This is what our Jam focused on. In today’s session, participants worked side by side with OpenAI mentors to find practical ways AI can support their daily work.Rather than starting from scratch, they explored building custom GPTs and reusable workflows they can apply in different situations—from situation reporting to needs assessment and public communication. The sessions also emphasized the importance of responsible use and building institutional trust in adopting AI technologies.

“This session is aimed at closing the gap between what AI can do and how it’s actually used in the field. Across Asia, there’s strong momentum and interest in AI, but the real opportunity is turning that into practical capability. By working directly with disaster-response professionals, we can ensure these tools are useful, accessible, and grounded in real-world needs.”

—Sandy Kunvatanagarn, Head of Public Policy at OpenAI

“Equipping the people closest to communities with the knowledge and skills to harness the power of digital tools and emerging technologies like AI is one of the most powerful investments we can make in disaster preparedness and response. We're proud to bring together partners across the region and to see it translate into tools that can be put to work right away.”

—Dr. Valerie Nkamgang Bemo, Deputy Director, Emergency Response at the Gates Foundation

“AI is opening new possibilities for how we understand and respond to disasters. ADPC integrates AI into geospatial tools and risk analytics to transform satellite and earth observation data into actionable insights. AI Skills Jam could improve AI literacy and empower people to find solutions to disaster challenges.

We can combine AI tools with regional expertise and partnerships to strengthen early warning systems, improve risk mapping, and support faster, more informed decision-making for communities and governments across the region.”

—Mr. Aslam Perwaiz, ADPC Executive Director

Together with our partners, we’re exploring a second phase in the coming months, focused on pilot deployments and deeper technical collaboration with participating organizations across the region. We look forward to continuing this work, building practical tools that help communities prepare for and respond to disasters more effectively. 

Generated by RSStT. The copyright belongs to the original author.

Source