Where Should International NGOs Start a Generative AI Journey?
A few weeks back, a conversation highlighted the current landscape of Generative AI for international NGOs (iNGOs). Many organizations are grappling with how to navigate this rapidly evolving technology, seeking clarity on what is practical versus what is merely hype.
Everyone Feels Behind on AI
Many in the NGO sector feel overwhelmed by the speed of AI development. Despite some donors creating grand AI strategies and organizations issuing press releases, the reality often reveals a landscape of cautious experimentation and internal confusion. Most organizations are still in the “Experimentation” phase, with many not even having reached that point, remaining stuck in a state of “Awareness”.
The current pace of change is unprecedented, outpacing previous tech revolutions. Unlike digital and mobile advancements that allowed for lengthy learning curves, AI demands rapid adaptation, with cycles of just 3-6 months. This places significant pressure on iNGOs to respond quickly to evolving public expectations and policy debates.
What Are Other iNGOs Doing?
Understanding the actual activities of other iNGOs amidst the noise is crucial. Activities can be categorized into five overlapping layers:
- Personal Use: Many teams are already using AI tools like ChatGPT and Claude for tasks such as drafting reports and analyzing data, often in an informal, ad-hoc manner.
- Institutional Use: Some organizations are intentionally integrating AI into internal processes like knowledge management and donor reporting, although these efforts are often led by individual teams rather than a coordinated strategy.
- Customer-Facing Use: A few NGOs are applying AI externally, such as through chatbots for information access or AI-driven diagnostics, though this raises ethical questions about bias and transparency.
- Governance: Effective management of AI use is still lacking. Few organizations have established clear operational guidance or engaged in meaningful conversations about ethical AI use.
- Ecosystem Level: This involves coalition-building and regulatory advocacy, which remains a slow-moving but necessary aspect for coherent alignment in the sector.
5 Key GenAI Challenges for iNGOs
As organizations navigate AI adoption, they face several challenges:
- Slow Systems vs. Fast Tech: Governance structures often lag behind AI’s rapid evolution, making it difficult for organizations to adapt policies effectively.
- Shadow AI: Much of the AI use is informal and unmonitored, leading to productivity gains and risks occurring outside formal strategies.
- Skills Gaps: Staff are often unprepared to make AI-related decisions, leading to a reliance on external vendors for guidance.
- Weak Data: Many iNGOs struggle with messy, fragmented data, which compromises the effectiveness of AI implementations.
- Real Risks of Harm: AI systems can reinforce bias and exclude marginalized groups, raising significant ethical concerns.
4 Suggestions for iNGO AI Leadership
To foster meaningful progress, iNGOs should consider the following suggestions:
- Turn Shadow AI into a Learning Opportunity: Acknowledge and support the informal use of AI among staff to surface valuable insights and identify risks.
- Get Your House in Order: Prioritize internal-facing AI initiatives to build competence and governance frameworks before engaging with external communities.
- Build Governance that Can Keep Up: Develop flexible, adaptive governance structures that can respond quickly to the fast-paced AI landscape.
- Upskill Everyone: Ensure all staff acquire basic AI literacy to contribute meaningfully to discussions about risk and governance.
Where Should iNGOs Start Their AI Journey?
While there is palpable interest in AI within the sector, uncertainty prevails regarding how to begin. Most organizations are already on their AI journey, even if informally. Embracing this uncertainty and fostering open conversations across teams can lead to thoughtful and progressive steps in integrating AI into their operations.
