Podcast Summary: Global Progress in the AI Era: How GiveDirectly is using AI to deliver cash faster

Podcast: This Week in Global Development
Date: March 23, 2026
Host: Katherine Chaney (Devex Senior Editor, Special Coverage)
Guest: Nick Allardyce (CEO, GiveDirectly)

Episode Overview

This episode explores how GiveDirectly, an innovative international NGO, is leveraging artificial intelligence (AI) to revolutionize humanitarian response—especially the delivery of emergency cash transfers. The conversation dives into the operational, ethical, and philosophical shifts AI brings to global development, featuring practical insights from GiveDirectly’s real-world pilots and their ambitious “moonshot” of delivering cash assistance worldwide within five days of a crisis. The discussion covers concrete examples from Bangladesh and Africa, operational changes inside GiveDirectly, the balance of speed versus precision, dignity in aid, and responsible AI risk management.

Key Discussion Points & Insights

1. How GiveDirectly Uses AI in Practice

• AI Across the Program
  GiveDirectly integrates AI into nearly every aspect of its humanitarian and development work, with particular emphasis on crisis response, aiming to deliver emergency cash within 5 days of any crisis globally.

  “We are using AI across almost everything we do… from flood forecasting models to recipient communications.” — Nick Allardyce, [02:10]

• Core Areas for AI Application

  • Identifying Vulnerable Communities: Using machine learning models, satellite imagery, and mobile phone data to swiftly pinpoint those most in need after a disaster.
  • Recipient Communication and Enrollment: AI helps process and aggregate massive data from beneficiaries, enabling swift communication, enrollment, and verification—even with feature phones.
  • Tracking Delivery: Monitoring cash transfers in real time to ensure support reaches communities.

  “Our goal is to get emergency cash within five days of any crisis to survivors… anywhere on Earth.” — Nick, [03:36]

2. Grounded Example: Bangladesh Floods & Anticipatory Cash

• Anticipatory Action for Floods
  GiveDirectly pre-enrolls communities in Bangladesh using AI-powered flood forecasting (developed with Google.org) to deliver cash before disasters strike, so recipients can take precautionary measures.

  “We partner with google.org and a bunch of kind of machine learning models to help us identify this... and if a flood forecasting trigger goes, we can actually get cash to those communities in advance.” — Nick, [05:45]

• RCT on the Value of Speed
  An ongoing randomized control trial (RCT) measures outcomes for cash transfers disbursed pre-crisis, immediately post-crisis, and on typical timelines, providing data on the impact of speed.

  "We're going to be able to see exactly how much value we can get by pushing that speed barrier." — Nick, [05:45]

3. Addressing Technological Barriers

• Mobile Penetration and Access
  While smartphone access may be limited, GiveDirectly designs systems for feature phones to broaden inclusion.

  "The first answer is: build solutions that work for feature phones, not for smartphones..." — Nick, [08:37]

• Positive Spillovers
  Cash transfers generate spillover benefits for communities, even those not directly reached, amplifying the program’s impact.

  "People who lived within half an hour of direct beneficiaries benefited almost as much as those who got the direct cash..." — Nick, [08:37]

4. Targeting Approach: False Positives vs. False Negatives

• Saturation Model vs. Precision Targeting
  GiveDirectly adopts a saturation approach, prioritizing inclusion—even if it means some people who are less in need also receive aid—to avoid missing those truly in need.

  “We try to reach everyone in them… we think it's worth it for a couple of reasons. One, it means that no one's left behind.” — Nick, [13:34]

• Efficiency and Dignity
  The inclusion-first approach reduces lengthy vetting and fieldwork, allowing resources to go directly to beneficiaries and preserving their dignity.

  “The incentives of both sides of that transaction are just like awful… from the recipient's kind of perspective and experience.” — Nick, [17:23]

5. Importance of Rigorous Evaluation

• RCTs and Measurement Culture
  GiveDirectly invests in RCTs to rigorously measure the impact of speed and anticipatory cash, sharing findings to influence the broader sector and governments.

  “We've since our founding had this huge emphasis on measuring our own work… and then also publishing that in public.” — Nick, [19:06]

6. Partnerships with Governments and Scaling the Model

• Government Collaboration
  There’s significant interest from governments in using AI-powered, rapid-disbursement systems, though technical capacity and infrastructure are often barriers.

  "We see a lot of hunger from governments… the barriers are often… technical capability or infrastructure." — Nick, [22:05]

• Infrastructural Vision
  The goal is to build a universal dashboard—a “portal” that NGOs or governments can use to reach people quickly with cash and track transfers transparently.

  “Think of it as a dashboard that is as simple as possible… and we can kind of see every step of the way how that money is flowing down to the community.” — Nick, [25:13]

7. Operational & Organizational Changes in the AI Era

• Talent & Team Structure
  GiveDirectly prioritizes hiring candidates with strong technical backgrounds, encourages staff-wide AI experimentation (e.g., hackathons), and enables quick adoption of new tools.

  “We've done org-wide hackathons… The only requirement is that you demo something at the end of the two days.” — Nick, [29:50]

8. Navigating Risks of AI in Aid

• Different Risk Profiles: Internal vs. Beneficiary-Facing AI
  Productivity tools for staff have less risk, but recipient-facing AI (e.g., targeting algorithms) demands strict privacy, testing, and ongoing community engagement.

  "Anything that involves supporting decision making around targeting of communities… we will rigorously test." — Nick, [33:19] “We are constantly running these focus groups with recipient communities to understand what’s important to them…” — Nick, [33:19]

• Dignity, Privacy & Trust
  Recipients voiced strong priorities for speed, fairness, and privacy, especially concerns about intra-community data visibility.

  “The value that those communities place on the saturation model... privacy within communities is obviously something that's very important to people as well.” — Nick, [33:19]

9. AI in Other Sectors: Malawi Agricultural Advice

• Pairing Cash with AI-Driven Advice
  In Malawi, GiveDirectly pilots programs combining cash transfers with AI-enabled agricultural advice, aiming to empower recipients not just with information, but with resources to act.

  “I think AI has the potential to democratize access to information and decision making… It’s only really going to create value if people have the resources.” — Nick, [37:29]

• Early Learnings
  Initial feedback shows that recipients value independent, unbiased advice, especially when community discussions can be stigmatizing or risky.

  "They fear being exploited, they fear being judged… having a neutral arbiter… in these pilots, recipients say that they really value." — Nick, [39:52]

10. Advice for the Sector

• Message to Other CEOs/Leaders
  Nick’s guidance is simple but emphatic:

  “Lean in. … This is really a kind of unimaginable opportunity for us to be more efficient, more effective, and support more people as a result. … We have to take advantage of these technological revolutions to reimagine what's possible.” — Nick, [40:50]

Notable Quotes & Memorable Moments

• On the Moonshot:

  “Our goal is to get emergency cash within five days of any crisis to survivors of crisis anywhere on Earth.” — Nick, [03:36]

• On Dignity:

  “…when you look at how vulnerable these communities are… the incentives of both sides of that transaction are so just like awful when you think about it…” — Nick, [17:23]

• On AI and Opportunity:

  “There's one world where AI creates unfathomable wealth… largely in the global north… and another world where we help communities get to kind of basic floors of opportunity…” — Nick, [37:29]

• On Risk and Privacy:

  “…privacy within communities is obviously something that's very important to people… The fact that we were kind of going to everyone in a community was like a significant mitigant…” — Nick, [33:19]

• On Organizational Culture:

  “…the most interesting stuff for the use of technology… is often going to come bottom up, not top down…” — Nick, [29:50]

• On Advice to Leadership:

  “Lean in… We have such a high responsibility to make sure every dollar goes as far as it can… we have to take advantage of these types of technological revolutions…” — Nick, [40:50]

Timestamps for Important Segments

• [02:10] – How GiveDirectly uses AI operationally
• [05:45] – Case study: Bangladesh flood forecasting and anticipatory cash
• [08:37] – Overcoming the technology access gap and positive spillovers
• [13:34] – Targeting strategy: inclusion, data use, and mobile phone data
• [17:23] – Dignity in aid: avoiding intrusive assessments
• [19:06] – The importance of RCTs and public evidence sharing
• [22:05] – Government partnerships and scaling the dashboard model
• [25:13] – Vision for infrastructure: the “dashboard” for crisis response
• [29:50] – Internal transformation: talent, hackathons, and tool adoption
• [33:19] – Managing AI risks: organizational principles and recipient focus groups
• [37:29] – AI in agriculture: combining advice and cash, feedback and learnings
• [40:50] – Advice for leaders: “Lean in” to AI opportunity

Key Takeaways

• AI is not theoretical in development aid—it is being operationalized now, changing how and how fast people in crisis receive help.
• Inclusion, dignity, and speed are prioritized over perfect precision in targeting, with AI unlocking new models of scalability.
• Rigorous impact measurement and partnership with governments are crucial for scaling and sector transformation.
• Empowering teams to experiment with AI and democratizing its benefits are seen as essential ingredients for both humanitarian effectiveness and organizational success.
• Leaders in global development should actively engage with AI’s opportunities and risks, or risk missing a transformative opportunity.