Shrinking humanitarian and development funding is putting increasing pressure on evaluation offices, which are expected to deliver more with fewer resources. But I also see an opportunity here: to rethink how we conduct evaluations and make better use of the evidence we already have.
In our joint blog, Doing More with Less: AI for Humanitarian, Environmental & Development Evaluation, we explored how AI could help evaluation offices respond to these pressures. For example, AI can support the synthesis of large volumes of documents, help identify evidence gaps, analyse evaluation portfolios and facilitate translation. Used responsibly, it could reduce time spent on some analytical tasks and allow evaluators to focus more on interpreting findings, understanding context and exercising professional judgement.
There is also an opportunity to look beyond efficiency gains in individual evaluations. Better use of existing evidence could help evaluation offices identify where new evaluations are most needed and avoid duplicating work where sufficient evidence already exists.
Of course, doing more with less cannot come at the expense of evaluation quality, independence or inclusiveness. AI-generated outputs need to be validated, and appropriate safeguards must be in place. Ultimately, the question is not simply whether we can produce evaluations faster or at lower cost, but whether we can use our resources more strategically to generate credible evidence that informs decisions and strengthens accountability and learning.
RE: Evaluation under pressure : Strengthening evidence use, accountability and learning amidst funding constraints (Joint EvalforEarth-ALNAP discussion)
Italy
Anoop sharma
Corporate Evaluation, AI and Data Specialist
IFAD
Posted on 13/10/2026
Shrinking humanitarian and development funding is putting increasing pressure on evaluation offices, which are expected to deliver more with fewer resources. But I also see an opportunity here: to rethink how we conduct evaluations and make better use of the evidence we already have.
In our joint blog, Doing More with Less: AI for Humanitarian, Environmental & Development Evaluation, we explored how AI could help evaluation offices respond to these pressures. For example, AI can support the synthesis of large volumes of documents, help identify evidence gaps, analyse evaluation portfolios and facilitate translation. Used responsibly, it could reduce time spent on some analytical tasks and allow evaluators to focus more on interpreting findings, understanding context and exercising professional judgement.
There is also an opportunity to look beyond efficiency gains in individual evaluations. Better use of existing evidence could help evaluation offices identify where new evaluations are most needed and avoid duplicating work where sufficient evidence already exists.
Of course, doing more with less cannot come at the expense of evaluation quality, independence or inclusiveness. AI-generated outputs need to be validated, and appropriate safeguards must be in place. Ultimately, the question is not simply whether we can produce evaluations faster or at lower cost, but whether we can use our resources more strategically to generate credible evidence that informs decisions and strengthens accountability and learning.