ICT4D Conference 2026 Nairobi: AI, Humanitarian Tech, and the Practitioner-Researcher Divide

The ICT4D Conference met in Nairobi in May 2026, running from the 20th to the 22nd in hybrid format — in-person in Kenya and accessible online for participants unable to travel. Founded by Catholic Relief Services in 2010 and now produced in partnership with TechChange, the conference has grown into one of the primary annual gatherings where NGO program staff, bilateral donor representatives, researchers, technology developers, and government officials working in digital development convene outside the formal academic circuit.

Its agenda this year — 85 sessions selected from more than 475 submitted abstracts — gives a reasonably legible picture of what the practitioner-adjacent ICT4D field is actually working on in 2026, as distinct from what academic conferences like IFIP WG 9.4 or ICTD are publishing.

The practitioner-researcher distinction matters. These are overlapping but not identical communities, and the gap between them has been a recurring topic of concern in both directions: researchers complaining that practitioners do not use evidence; practitioners observing that academic research arrives too late, is not actionable, and does not address the specific operational questions that matter in real programs.

AI Dominates the Agenda

By the conference organizers’ own account, artificial intelligence dominated both the abstract submissions and the final program. That is consistent with what the broader sector has been experiencing since 2023: a flood of AI-enabled tools and vendor pitches directed at humanitarian and development organizations, alongside a growing and legitimate concern about whether these tools work, for whom, and under what conditions.

The AI sessions at the 2026 ICT4D conference spanned a wide range: AI for data analysis in monitoring and evaluation, AI-enabled chatbots for beneficiary communication, predictive analytics for targeting cash transfer programs, AI applications in agricultural advisory services, and AI governance frameworks for humanitarian organizations.

This breadth reflects a sector-wide pattern. Development and humanitarian organizations are not waiting for the research evidence on AI to mature before deploying AI tools — procurement cycles, donor interest, and the availability of funded “innovation” programs mean that deployment often precedes evaluation. The conference’s AI sessions were partly a documentation of what is already happening.

What was less visible at the conference, based on available program summaries, was rigorous outcome evidence — evaluations showing that AI applications in development contexts produce the development outcomes they are supposed to produce. The systematic review literature on AI4D, still thin, does not yet provide confident generalizations about what works.

Responsible Data: The Recurring Problem

A second major theme across the 2026 Nairobi agenda was data responsibility — covering responsible data sharing, interoperability, and the power dynamics embedded in data collection and use.

This is not a new conference theme. Responsible data has been on the ICT4D conference agenda in various forms since at least 2016, when the humanitarian sector’s exposure to data misuse in conflict contexts became visible through several high-profile cases. What changes in 2026 is the data environment being discussed: much larger datasets, more pervasive collection through digital ID systems and social protection registries, and the AI processing that makes these datasets newly consequential.

The power dynamics dimension is significant. When a bilateral donor funds a digital social protection program and requires beneficiary biometric data as a condition of funding, questions about who owns that data, who can access it, and what happens to it after the program ends are not primarily technical questions. They are governance questions about the relationship between donors, governments, and populations receiving assistance.

ICT4D researchers have documented these dynamics in specific country contexts — but the broader sector frameworks for governing data in development programs remain contested and unevenly implemented.

The Practitioner-Researcher Divide

The 2026 Nairobi conference explicitly describes itself as a gathering of NGOs, local organizations, governments, donors, researchers, and private sector partners. In practice, conference attendance and program composition skew toward the practitioner end of that spectrum.

This has implications for what knowledge gets validated at the conference. The abstract selection process for a practitioner-dominated conference will weight relevance to current programming, accessibility of presentation, and connection to active programs more heavily than methodological rigor or theoretical contribution. That is entirely appropriate for the conference’s purpose — but it means the knowledge produced and validated at Nairobi is different from, and not always cumulative with, the knowledge produced at academic ICT4D venues.

The productive tension between these two knowledge-production contexts has not been resolved. Researchers who want their work to influence practice need to engage with practitioner forums; practitioners who want to avoid repeating documented mistakes need access to research findings in formats they can use. The ICT4D Conference in Nairobi is one of the few spaces where both communities are present simultaneously, which makes it a genuine — if imperfect — site for dialogue.

Several research institutions now use the conference explicitly as a dissemination venue: IFPRI, the International Food Policy Research Institute, participated in the 2026 conference as a named institutional partner, presenting research on agricultural technology and food security. J-PAL (the Abdul Latif Jameel Poverty Action Lab) and IPA (Innovations for Poverty Action) have historically used practitioner conferences as a space to communicate impact evaluation findings to implementing organizations.

What Nairobi and Kathmandu Both Reveal

Running in the same two-week window in May 2026, the ICT4D Conference in Nairobi and the IFIP WG 9.4 conference in Kathmandu together constitute a reasonable cross-section of where the digital development field is intellectually.

Nairobi’s agenda shows the operational pressures: AI is being deployed faster than it is being evaluated; data governance frameworks are inadequate to the systems being built; practitioners need tools for navigating these realities now, not after five more years of academic research.

Kathmandu’s theme shows the foundational questions: the frameworks being used to design, evaluate, and govern ICT4D programs are themselves shaped by assumptions that deserve critical scrutiny — about what development is, who gets to define it, and whose knowledge counts in the design of digital systems.

Neither conference can substitute for the other. The ICT4D field needs both the operational grounding of Nairobi and the critical reflexivity of Kathmandu — and it needs better mechanisms for knowledge to move between them.

The practitioner-researcher divide will not be bridged by better conference planning alone. It requires structural changes in how development research is funded, how careers in research are rewarded, and how implementing organizations build their own analytical capacity. These are the long-term questions that neither May conference fully answered — but both, in their different registers, put on the table.

Further reading from authoritative sources:

• ICT4D Conference 2026 — ict4dconference.org — official conference site with program details, session recordings, and speaker information.
• IFPRI @ ICT4D Conference 2026 — International Food Policy Research Institute’s participation, with context on agricultural technology and food security research presented at the conference.