Humanitarian Frontiers – "The Donor Dilemma: Risk Tolerance, Innovation and Responsibility" Podcast Summary

Main Theme and Purpose This episode explores the shifting landscape for donors funding AI-driven innovation in the humanitarian sector amid a challenging financial climate. Host Chris Hoffman is joined by a panel of influential donors and innovation leaders: Teresa Marie Upstrom Pankhatov (Humanitarian Innovation Program, Innovation Norway), Sean White (UK Humanitarian Innovation Hub), and Zaina Alsaman (Creating Hope in Conflict, Grand Challenges Canada). Together, they dissect the delicate balance between fostering innovation, managing risk, engaging affected communities, and ensuring responsible data stewardship as new technologies and funding models redefine the boundaries of humanitarian aid.

Key Discussion Points & Insights

1. Sectoral Flux and the Funding Crunch

• Setting the Stage (02:51): Chris introduces the guests and the central concern: how are donors adapting their priorities in an era marked by drastic funding cuts and shifting expectations for innovation?

Sean White (UK Humanitarian Innovation Hub)

• Notes the "grim time" for sector funding, urging reflection: "Are we just fixing and dabbling and tweaking the branches, or are we really dealing with some of the root causes of this fragility that we face at the moment?" (03:15)
• Emphasizes the assumed efficiency gains from AI but highlights a lack of robust evidence: "Do we really have the evidence at the moment that AI development is going to be cost saving?" (04:10)
• Calls for reducing competition and duplication, promoting coordinated investment, and ending "projects that simply duplicate what's already being developed" (05:20).

2. Innovation Beyond Life-Saving: Importance of Systems Change

Teresa Marie Upstrom Pankhatov (Innovation Norway)

• Celebrates Norway's continued funding and stresses support for innovations beyond just 'life-saving' contexts: "It's not just about life-saving, it's about innovating across the whole process—procurement, greening operations, partnerships." (07:45)
• Stresses the need for systems innovation: "Any innovation with large potential pushes on the system and requires adaptation across processes and tools." (08:30)
• Highlights that partnerships across sectors are increasingly vital as resources tighten.

3. The Challenge of Scaling AI in Humanitarian Contexts

Zaina Alsaman (Grand Challenges Canada)

• Warns against AI as a "shiny, sexy thing" and stresses demand-driven solutions: "There’s a tendency of AI being this...catch-all to solve all the world’s problems. But we need to think about end-user demand at the community level." (11:12)
• Identifies the lack of local actors in AI-led projects as a barrier to effective scaling. "Most AI projects...are actually led by organizations in the US, Canada, and Europe...what we need to see...is more investment in local institutions." (12:35)

4. Risk Appetite and Responsible AI: Finding the Balance

Sean White

• Observes an evolving debate between "responsible AI" and "innovative AI": "According to me, there's a false dichotomy between these two things. You don't have to be either innovative or do responsible AI, you can do both." (15:10)
• Warns of "FOMO" leading to rushed, under-thought AI adoption.
• Stresses the importance of slow, needs-based innovation and involving frontline populations: "If things are explained well...even with low literacy, you can engage people." (17:45)

Zaina Alsaman

• Underscores “do no digital harm”: "As humanitarians...we need to ensure that we are being responsible data custodians...That extends to not only principles of do no harm, but...do no digital harm." (19:05)
• Stresses demonstrated risk mitigation in proposals: "Deeper due diligence in this area is really critical." (20:55)

5. Innovative Partnerships & Financing Models

Teresa Marie Upstrom Pankhatov

• Describes Innovation Norway’s 50/50 co-financing model: "If you want to scale an innovation, we require you to match our financing with financing from the private sector...It’s also an element of quality assurance—in that we're not the only one who'd like to invest." (23:53)
• Highlights the need for "informed consent" from communities about their data: "I keep coming back to...how can we have informed consent with people affected by crisis on how their data is collected...and the risks associated with it?" (25:05)
• Calls attention to the urgency of bridging the digital divide to ensure inclusive, unbiased AI.

6. Data Ownership, Privacy, and Regulatory Complexities

• Data Dilemmas: There is both scarcity and overabundance of data. Teresa and Naseem note affected populations may want compensation for their data. But as Sean acknowledges, "There is an AI data dilemma at a global level...crises will destabilize or make data irrelevant quite quickly." (30:10)
• Governance: Sean outlines three governance layers needed: organizational, sectoral, and global, noting "it’s not one mechanism we need...it’s a range of measures.” (34:07)
• Regulatory Patchwork: The panel discusses the impact of the EU’s new AI regulations and GDPR—even for organizations based outside Europe, compliance may become necessary when partnering across borders.
• Power Imbalances: Academic research and pilot projects are still dominated by actors from the Global North, risking “repeating our colonial legacies in the domain of AI.” (36:50)

7. Cultivating True Partnerships: Social Impact, Tech, Local Actors

• Chris: "There are...AI companies solely focused on working in the humanitarian sector...but the humanitarian sector [is] missing out in the conversations...the creation of partnerships, co-developing programs and co-seeking funds, is vital." (38:02)
• Zaina: Partnerships must extend to the community level. "There needs to be not only partnership with the private sector...but also partnership at the local community level." (40:19)

8. From Needs Assessment to Impact: The Role of AI

• Teresa: "We invest in needs, we don’t invest in solutions." Humanitarian actors must start with a clear needs assessment, then invite market dialogue—educating both sides before selecting tech partners. Sustainable partnerships are crucial: “If a partnership is not sustainable for both parties, it's not going to last.” (44:55)
• Intellectual property rights and licensing must be navigated carefully to allow both humanitarian principles and private sector involvement.

9. Real-World AI Use Cases and Their Risks

Notable Examples:

• Safe Optimization Tool (SWOT): Uses machine learning to optimize water chlorination in displacement settings by leveraging existing field data. (47:20)
• HOLA Systems: ML-based early warning for airstrikes, giving 5–7 minutes’ advance notice, credited with saving lives in NW Syria. (48:00)
• Flood Warning Systems/Chatbots: Multiple, sometimes duplicative efforts, highlighting both opportunity and risk for more centralized solutions. Sean on chatbots: “Most of us try and break the chatbot...we need to be very conscious of what type of service we are trying to provide.” (51:20)

Sean’s broader point: "There are actually a lot of...unsexy but really important back use cases that could help us navigate some of the bureaucracy we have built within the humanitarian system...If we are improving, if we are finding AI uses to alleviate some of the hours and hours that we spend on report writing or...backend tasks, how can we then get better at the stuff that's actually really important, like being a good listener to an affected population?" (52:25)

Notable Quotes & Memorable Moments

• Sean White on Funding Realities (02:52): "Let's face it, it's a really grim time...it's pretty hard at the moment to see pinnacles of hope amid these pretty drastic funding cuts across bilateral donors."
• Teresa Upstrom Pankhatov on Systems Change (08:30): "Any innovation…often pushes on the system and requires the system to change or adapt to be able to reach the impact that it can reach. And I think AI does that."
• Zaina Alsaman on Local Leadership (12:35): "What we need to see moving forward to ensure that these solutions are scalable is more investment in local institutions that are spearheading AI investments."
• Sean White on False Dichotomies (15:10): "There's a false dichotomy between responsible AI and innovative AI. You don't have to be either innovative or do responsible AI, you can do both."
• Zaina Alsaman on Data Responsibility (19:05): "We need to ensure that we are being responsible data custodians…not only do no harm, but do no digital harm."
• Teresa Upstrom Pankhatov on Informed Consent (25:05): "I keep coming back to...how can we have informed consent with people affected by crisis on how their data is collected...and the risks associated with it?"
• Chris Hoffman on Competition versus Collaboration (38:02): "We've always been a very competitive small group of social impact startups and companies...the creation of partnerships, co-developing programs and co-seeking funds...is vital."

Timestamps for Important Segments

| Segment | Speaker(s) | Timestamp | |-----------------------------------------|------------------------------|------------| | Donor perspectives on funding shifts | Panel | 02:51–06:55| | Systems innovation & partnerships | Teresa | 07:28–09:23| | Scaling challenges in AI for aid | Zaina | 10:27–13:14| | Risk appetite and responsible AI | Sean, Zaina | 14:26–21:33| | Balancing innovation and risk | Teresa | 23:38–28:21| | Data ownership & informed consent | Teresa, Sean | 28:24–32:10| | Regulatory/gov. challenges in AI | Sean, Zaina | 33:47–37:31| | Partnerships & leveraging expertise | Chris, Zaina | 37:31–41:38| | Needs assessment & sustainability | Teresa | 42:40–46:14| | Notable AI-driven use cases | Zaina, Chris, Sean | 46:17–53:54|

Tone and Takeaways

This episode is candid yet optimistic. Donors are acutely aware of hard realities—resource constraints, uneven power, underfunded innovation—but they remain committed to responsible progress. Central tenets include meaningful partnerships (especially with local actors), embedding responsible data practices, rigorous needs assessment, and investing boldly but carefully in technology as a change agent. The conversation is rich with examples, warnings, and a call to break with “business as usual” in humanitarian innovation.

In the words of Chris Hoffman (Closing, 54:47):
"This is what you get when you bring a whole bunch of friends together that love a subject and want to chat...This was really phenomenal. Thank you all for joining."

For listeners:
This episode offers a nuanced, inside look at the donor side of humanitarian innovation—its hopes, fears, contradictions, and the collaborative spirit required to move towards scalable, ethical, and effective AI solutions for those most in need.