With information scattered across multiple reports, public administrations often struggle to truly harness the value of evaluative evidence for decision making, accountability and learning. To make a more effective use of evaluation reports, Finland’s Ministry for Foreign Affairs developed OpenEval, an Artificial Intelligence (AI)-assisted platform that enhances information use across Finland’s administration and stakeholders, supporting decision making, transparency and organisational learning. The project emphasised internal capacity building, stakeholder feedback, and an adaptive development.
Using AI to make the most of evaluation evidence in Finland
Abstract
Public administrations often deal with vast amounts of precious evaluative evidence, yet making effective use of that information is challenging: the Finnish Ministry for Foreign Affairs (MFA) recognised that the findings of regular and systematic evaluations were scattered across a large number of documents, and textual evidence often underused. This limited the ability of users to extract insights from evaluation reports for decision making, learning and accountability. To address this, in 2025, the Ministry’s independent Development Evaluation Unit developed OpenEval, an AI-assisted platform designed to capture, organise and make accessible to both the public and development actors the evidence embedded in their reports.
Finland approached the leveraging AI technologies for more effective policy making in several stages:
Recognising the potential of AI. The MFA followed international developments in AI that supported enhanced access to evidence. It trained its staff and piloted some uses of data science and AI techniques in evaluation assignments, such as using Natural Language Processing to evaluate the human rights-based approach in Finland’s development policy and co-operation.
Leveraging institutional support. The AI project was aligned with, and supported by, the aspirations of the MFA and the broader Finnish government to advance digitalisation and the use of AI in public administrations.
Identifying easy entry points for AI piloting. Working with information from public evaluation reports was an easy entry point to pilot AI while limiting its risks. Additionally, using existing tender frameworks, with the relevant companies’ capabilities already validated, simplified the process.
Collaborating and including various fields of expertise internally and externally. A project steering group, including the project lead, head of unit, partner company representatives, and MFA's information technology (IT) staff, monitored progress. The steering group shared lessons with the MFA’s newly established AI task force. Feedback was also collected from external stakeholders such as evaluators.
Sequencing the iterative approach and ensuring viability. The MFA initially ordered three proofs-of-concept to check OpenEval’s viability and basic design elements. Following the start-up phase, a first version of the tool was developed. Three rounds of testing and feedback were then conducted with both internal and external stakeholders. Overall, the process took about nine months from start to launch. OpenEval is updated as new evaluations become available and is expected to be further developed based on user feedback.
Guiding staff use. To support the roll-out and uptake of OpenEval, an internal guidance framework “Reuse, refine and repurpose” outlines the different ways the tool can serve staff in providing raw materials and inputs for various planning and administrative tasks. Technical training and promotion of use have encouraged the adoption of this new technology.
A pioneer project resulting in a “public good”. OpenEval is the MFA's first open-access AI-assisted tool serving civil servants, stakeholders and the general public. It has significantly advanced the MFA's AI development. This project has spurred the establishment of the necessary IT architecture and processes for future similar initiatives.
More effective use of evidence. OpenEval allows users to search through multiple reports simultaneously and quickly, reducing time and effort. It enables better use of the extensive information available on Finland’s development co-operation and related issues such as support for human rights. For example, OpenEval serves officials who prepare projects, analyses and communication materials.
Efficiency gains. OpenEval is designed to assist, not replace, staff in their duties. It brings efficiency gains in processing information for forthcoming evaluation and review assignments. For example, after its launch, MFA staff shifted from manually extracting policy-relevant evaluative evidence from evaluation reports to an automated information management system. The tool performs data search and extraction by key words in seconds. An initial estimate indicates that searching and extracting 100 search results manually takes about 20 minutes by an advanced user. The larger the corpus, the larger the efficiency gains.
Transparency and evidence-based decision making. By improving the accessibility of evidence on development policy and co-operation to the public and stakeholders alike, OpenEval supports transparency in governance, organisational learning and evidence-based decision making. It complements other transparency tools and sources of information, such as Finland’s OpenAid website. The tool provides a detailed overview of evidence coverage and highlights the need for new evaluative information where evidence gaps exist.
Define clear objectives. Before starting, the MFA defined what it wanted to develop, why, and what was expected from the AI tool. An iterative and adaptive process was necessary, but a clear vision of the objective and required functionalities was key to designing a tool that served its purpose.
Design a process that builds internal capabilities. A participatory approach –involving both development co-operation staff and IT specialists– and lesson sharing were beneficial throughout the process. Basic training sessions on AI-related topics helped build a shared sense of direction among staff and facilitated their participation in the process.
Leverage existing partnerships and frameworks. If the required in-house capabilities are lacking, contracting external partners can be a solution. Tapping into existing tenders or framework arrangements can simplify the process.
Ensure participation and promote feedback. Sharing key project milestones with internal and external stakeholders feeds into the development of the AI tool and improves it. Opportunities for engagement span from testing AI tools within ongoing evaluation assignments to raising awareness of AI use cases during events.
Dedicate sufficient time and resources. The process of developing OpenEval took about nine months from start to launch. It required investment in time and commitment by the project lead and other staff members for ongoing consultation and testing of the tool at various stages.
Ministry for Foreign Affairs of Finland, Development Evaluation, https://um.fi/evaluation-of-development-cooperation.
Finland’s OpenAid website, https://openaid.fi/en/.
OpenEval, https://www.openeval.fi.
OECD (2024), OECD Development Co-operation Peer Reviews: Finland 2024, OECD Development Co-operation Peer Reviews, OECD Publishing, Paris, https://doi.org/10.1787/2a9a43a2-en.
OECD (2024), Recommendation of the Council on Artificial Intelligence, C/MIN(2024)16/FINAL, https://legalinstruments.oecd.org/en/instruments/oecd-legal-0449.
Lorenz, P., K. Perset and J. Berryhill (2023), “Initial policy considerations for generative artificial intelligence”, OECD Artificial Intelligence Papers, No. 1, OECD Publishing, Paris, https://doi.org/10.1787/fae2d1e6-en.
OECD (2021), Applying Evaluation Criteria Thoughtfully, OECD Publishing, Paris, https://doi.org/10.1787/543e84ed-en.
To learn more about Finland’s development co-operation, see:
OECD, “Finland”, Development Co-operation Profiles, OECD Publishing, Paris, https://read.oecd.org/10.1787/3fffb451-en.
More In Practice examples from Finland are available on Development Co-operation TIPs • Tools Insights Practices.
This work is published under the responsibility of the Secretary-General of the OECD. The opinions expressed and arguments employed herein do not necessarily reflect the official views of the Member countries of the OECD.
This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area.
Photo credits: ©NicoElNino/Shutterstock.
© OECD 2025
Attribution 4.0 International (CC BY 4.0)
This work is made available under the Creative Commons Attribution 4.0 International licence. By using this work, you accept to be bound by the terms of this licence (https://creativecommons.org/licenses/by/4.0/).
Attribution – you must cite the work.
Translations – you must cite the original work, identify changes to the original and add the following text: In the event of any discrepancy between the original work and the translation, only the text of original work should be considered valid.
Adaptations – you must cite the original work and add the following text: This is an adaptation of an original work by the OECD. The opinions expressed and arguments employed in this adaptation should not be reported as representing the official views of the OECD or of its Member countries.
Third-party material – the licence does not apply to third-party material in the work. If using such material, you are responsible for obtaining permission from the third party and for any claims of infringement.
You must not use the OECD logo, visual identity or cover image without express permission or suggest the OECD endorses your use of the work.
Any dispute arising under this licence shall be settled by arbitration in accordance with the Permanent Court of Arbitration (PCA) Arbitration Rules 2012. The seat of arbitration shall be Paris (France). The number of arbitrators shall be one.
