ACUTE project on AI in COI: Report on technological approaches
The ACUTE Project has released another deliverable, presenting an overview on AI technologies relevant for COI research.
We are pleased to present another publication from the ACUTE project: the report “Enhancing Country of Origin Information Research with AI: Relevant Technological Approaches and Solutions”. This report provides an extensive overview of current technological developments and approaches for the use of artificial intelligence (AI) in the COI context.
The report was prepared by the Hub of Computing & Data Science of the University of Hamburg, commissioned by ACCORD for the ACUTE project.
"The objective of this report is to document and present the possibilities of artificial intelligence (AI) and the technologies that enable and underpin it, often described as machine learning (ML). The report presents an overview of technologies to help readers understand the possibilities, fundamental differences between existing technical approaches and requirements, and the risks of training ML models and using them. This report focuses on three main areas. First, a foundation is laid out by explaining AI terms and fundamentals (cf. Section 2). Second, the report presents technological approaches such as supervised and unsupervised learning, neural networks, and the transformer architecture, which lay the foundations for large language models (cf. Section 3). Third, we describe and summarize possibilities and challenges in using large language models, including extensions to make them domain-specific (cf. Section 4). Finally, we discuss the findings in the context of the country-of-origin domain and derive recommendations towards a responsible use of AI systems in this critical and sensitive domain (cf. Section 5)." (p. 1)
Building on this structure, the report provides a detailed, technically grounded overview of key developments in the field of machine learning and discusses their relevance for COI research.
At the same time, it discusses the limitations and risks associated with these technologies. In particular, it addresses challenges related to data quality, bias, hallucinations and security risks such as prompt injection.
The report is intended as a technical and conceptual foundation for further work within the ACUTE project. It complements earlier outputs on needs and requirements and contributes to preparing the ground for the development of concrete applications and guidelines.
- Hub of Computing & Data Science, University of Hamburg: Deliverable 2.2: Enhancing Country of Origin Information Research with AI: Relevant Technological Approaches and Solutions, June 2026
https://www.ecoi.net/site/assets/files/7967/d2-2_enhancing_country_of_origin_information_research_with_ai_2026-06.pdf
The ACUTE project aim to contribute to a common approach and help shape practical guidance for the responsible use of AI in the field of COI.
We invite readers interested in COI, AI, and responsible innovation to explore the publication and follow the project’s ongoing work in the years ahead. You can read a project description on ACUTE here.
