Humanitarian AI Today

Priyank Hirani, Vice President of Programs at data.org discusses the India AI Impact Summit and India’s role as a development space and test bed for responsible and inclusive AI applications with Humanitarian AI Today Producer, Brent Phillips.

In this insightful conversation, Hirani highlights how India is shifting the global narrative away from frontier model competition toward practical, population-scale AI deployment that addresses real-world challenges in health, agriculture, and finance. By building directly upon a decade of robust digital public infrastructure, the Indian ecosystem accelerates adoption through lean, cost-effective and context-aware innovation. Hirani emphasizes how this unique ecosystem serves as a powerful blueprint for the global majority, championing "South-South" collaborations and responsible and inclusive AI innovation.

The episode underscores data.org’s mission as a global catalyst, capacity builder, and convener in the social impact sector. Hirani discusses the organization’s ambitious goal to cultivate one million purpose-driven data professionals by 2032 through its expanding Capacity Accelerator Network. Looking ahead to the future of technology, he shares a compelling vision for "Civic AI” and the creation of persistent, multilingual, and context-aware personal AI agents designed to help citizens frictionlessly navigate complex public systems. The conversation is packed with updates on data.org’s latest initiatives, including regional hackathons and the Finverse, a decision support tool and resource hub designed to help organizations in the Asia Pacific (APAC) region use data and AI to improve the financial health of the individuals and communities they serve. This episode is a must-listen for social impact professionals, researchers, and anyone invested in a more equitable and human-centered AI future.

In this Humanitarian AI Today Voices flashpod, Eric Talbert, Co-founder of MedCycle Network guest hosts an interview with Jarrod Goentzel, founder and director of the MIT Humanitarian Supply Chain Lab in the MIT Center for Transportation & Logistics.

This interview dives into the evolution and modern practices of the MIT Humanitarian Supply Chain Lab for humanitarian professionals looking to optimize crisis response through system-level thinking and technology. The discussion traces the lab’s journey from its origins during the 2004 Indian Ocean tsunami to its needs-assessment work and market-resilience studies to it’s general shift away from reactive, event-specific planning toward building structural, "system-level" understandings of supply chains and how organizations can better anticipate bottlenecks and coordinate with the private sector.

For humanitarian professionals, the interview offers a grounded, pragmatic perspective on integrating artificial intelligence into crisis response. Goentzel explicitly addresses the limitations of relying solely on automated systems, noting that AI inherently struggles with data gaps, as it is bounded by what is publicly available and cannot easily synthesize entirely unique disaster contexts on its own. To overcome this, the MIT Humanitarian Supply Chain Lab utilizes AI as an initial data-gathering and pattern-matching catalyst, which is then verified through a human-in-the-loop framework. The lab deploys a network of real-time ground-truthers who are trusted professionals embedded within the supply chain who validate the AI's outputs. This hybrid model ensures that automated data collection never compromises the absolute operational integrity required when delivering life-saving aid to vulnerable populations.

The interview touches upon "polycentric governance," which is the concept of humans organically cooperating to manage common resources during crises. The lab models supply chains as complex adaptive systems and conducts "Blue Sky Studies”which are highly detailed structural mapping done when there is no active emergency to locate vulnerabilities before disaster strikes. A prime example of this is the lab’s SCAN (Supply Chain Analysis Network) mapping, which evaluated infrastructural bottlenecks in transportation and fuel pipelines.

Looking toward the future of humanitarian tech, the conversation highlights cutting-edge applications of predictive modeling and advanced AI training. For AI developers, Goentzel offer’s a futuristic vision for disaster AI: rather than letting a machine application start from scratch during an active crisis, the lab is actively researching ways to pre-embed AI with complex supply chain network science and system dynamics. By providing the machine with a sophisticated baseline of structural interdependencies beforehand, the AI can immediately interpret real-time news and data influxes with extreme speed. This effectively frees up human humanitarian leaders to step away from the information onslaught and focus entirely on creating the rapid physical and collaborative connections needed to save lives.

The MIT Humanitarian Supply Chain Lab offers resources and educational platforms to connect researchers, technology experts, and ground-level aid workers. Goentzel invites listeners to join the lab’s humanitarian supply chain community and take advantage of free online course developed by the lab, like the lab’s free Humanitarian Logistics course through MITx: https://www.edx.org/learn/business-administration/massachusetts-institute-of-technology-humanitarian-logistics

An article on the Lab's supply chain resilience work can be found here: https://ctl.mit.edu/sites/default/files/documents/scmr-innovation-strategies-september-2025.pdf

To learn more about Eric Talbert’s work and the MedCycle Network, check out his interview on the Grow Healthy, Help People Podcast: https://youtu.be/w495cOVVajw?si=EMZLr-zZXAWM93Oq

Rubén Lozano-Aguilera, Product Lead for Asta at the Allen Institute for AI (Ai2), explores the transformative potential of agentic AI in scientific research with Humanitarian AI Today guest host Lindsey Moore, Founder of DevelopMetrics. Rubén introduces AutoDiscovery, a powerful new tool developed by his team that moves beyond traditional query-based analysis to autonomously generate and test scientific hypotheses. This shift from manual data processing to autonomous discovery offers a powerful force multiplier for researchers, helping them surface blind spots and hidden patterns that traditional methods often overlook in fields ranging from melanoma research to marine ecology.
 
For humanitarian and development organizations, Ai2's work represents a vital new advancement in what Rubén calls "shared AI infrastructure." Ai2's deep commitment to the open-source movement, providing open models, checkpoints, code, and training data, ensures transparency and accessibility for all. This approach is particularly impactful for organizations operating in resource-constrained environments, as it allows them to leverage state-of-the-art predictive analytics without the high costs or "black box" risks associated with proprietary systems. By democratizing access to high-level research tools, Ai2 enables any researcher or developer to maintain data ownership while utilizing sophisticated AI to solve the world's most pressing problems.
 
The conversation next turned to the deeper philosophical stakes of automating scientific discovery itself. Drawing on his graduate research in AI ethics at Cambridge, Rubén separates what philosophers of science call the "context of discovery”, how a hypothesis is generated, from the "context of justification," how it is tested and validated. The worry is deskilling: as scientists offload hypothesis generation to AI, will they lose the instincts needed to catch when the machine is wrong? His answer centers on cultivating "meta-AI skills", the practiced ability to evaluate AI outputs critically. That raises its own problems: how do those skills get built, and are they really the same kind of skill as the hypothesis-generating instincts they would replace? Ai2 is actively studying this by examining how tools like AutoDiscovery affect students and early-career researchers. For humanitarian and development professionals navigating an era of shrinking research budgets and growing AI adoption, these added points raise essential questions about keeping human judgment at the center of discovery.

Ethna Ghosh, India Partnerships Lead with Renaissance Philanthropy, discusses the nonprofit organization’s operational strategy, her work and India’s role as a global hub for AI experimentation with Brent Phillips, Humanitarian AI Today podcast producer.

Drawing on her deep roots in the Indian development sector, Ethna shares insights from her experience as a founding member of GivingPi, India's first family philanthropy network at Dasra, where she specialized in building strategic partnerships and advising families on high-impact giving. From this vantage point, she provides an insider’s perspective on India’s emergence as a science and technology pioneer and a premier real-world testing ground, where localized AI applications can be deployed across a diverse population of 1.4 billion people with immense cultural, geographic and linguistic diversity to address a myriad of social challenges.

Elaborating on her role as Renaissance Philanthropy’s India Partnerships Lead, Ethna outlines how the organization accelerates scientific, technological and innovation breakthroughs by using a thesis-driven, time-bound fund model connecting donors with domain experts to tackle high-impact, underfunded projects. By empowering field leaders to identify and eliminate systemic bottlenecks to enable scalable change, the organizations acts as force multiplier to achieve large-scale, transformative impact rather than just deploying capital. By supporting a diverse portfolio of deep-science innovations, the organization seeks to spark a modern "Renaissance movement" where converging technologies like AI and space tech for example transform the global landscape.

Offering insight into the keys to success in environments like India, Ethna emphasizes "bottom-up" approaches that prioritize deep collaboration between Samaj (society), Sarkar (government), and Bazar (market). Highlighting rural maternal health workers as a case study, Ethna points out that involving government entities is a non-negotiable in India for any initiative aiming to achieve true national scale and effectively engage the "last mile" of the population.

Looking toward the future, Ghosh envisions a shift toward agentic and physical AI that moves beyond simple text prompts to become a voice-enabled "buddy" for those with limited digital literacy. She advocates for the development of inclusive models that strip away linguistic and gender biases to provide safe, life-changing access to justice, healthcare, and government services. This vision ensures that technology serves as a bridge for equality, providing every citizen with a reliable companion to navigate a rapidly changing world. She concludes with a powerful call for radical collaboration among foundations and NGOs, urging stakeholders to move past silos and genuinely partner to eliminate the bottlenecks holding back progress.

Co-produce by Humanitarian AI Today, this third pilot episode of The Inference Layer podcast bridges the technical complexities of AI deployment with the reality of humanitarian operations and dives into the transition from static models to autonomous agentic systems. On behalf of the Humanitarian AI Today podcast, guest host Patrick Hassan, an AI policy lead with a background in disaster response, interviews Federico Pierucci, Scientific Director of the Icaro Lab, to explore how the inference layer is becoming a site of significant systemic risk. The discussion provides a unique look at inference-time failures such as alignment drift and steganographic coordination that emerge only when multiple agents interact in production environments.

For humanitarian actors, the episode raises concerns regarding operating in an era of assistance automated by layers of AI agents. The dialogue highlights how multi-agent chains used for beneficiary selection or resource allocation for example can degrade, develop invisible biases or be weaponized or politicized by parties to a conflict. Federico explains that these risks can be compounded by a lack of safety benchmarks for things like underrepresented languages and dialects, which can lead to unpredictable jailbreaks or administrative failures in the field.

The episode provides an inside look at pioneering research being carried out by the Icaro Lab, a Rome-based laboratory specialized in AI safety in collaboration with the Sapienza University. The lab focuses on mechanistic interpretability, a technical field dedicated to understanding the internal attention heads and decision-making units of an AI to decipher how it truly processes information. The discussion introduces the concept of Institutional AI, a proposed framework to manage these emerging xeno-behaviors through a governance graph. Rather than relying solely on prompt engineering or model-level alignment, Federico argues for a protocol-level solution that can manage misbehaving agents during inference. The episode is informative for professionals seeking to understand why AI safety must evolve from a localized technical challenge into a global institutional design problem, particularly in regions where traditional governance has broken down.

This particular episode moves beyond surface-level AI ethics and safety issues that the humanitarian community has been talking a lot about, to address inference-time vulnerabilities in agentic systems. This is an important topic because as the humanitarian community moves from developing and testing simple chatbots to incorporating autonomous multi-agent systems into humanitarian operations, we face new challenges that can have very serious consequences - making the 'inference layer' a new frontier for humanitarian risk.