Generative AI startup Fortytwo Maru said on Tuesday it is stepping up the use of AI for green finance with BNZ Partners through an “AI-based K-Taxonomy compliance assessment support system (BNZ AITM)”.

The project combines BNZ Partners' K-Taxonomy expertise with Fortytwo Maru's generative AI technology. It aims to support fast, accurate and consistent application of K-Taxonomy classification needed in the ESG certification application process.

Fortytwo Maru handled the entire process, from automating business summaries and K-Taxonomy classification to Q&A and installation and training. It combined its domain-specific lightweight model LLM42 with retrieval-augmented generation RAG42 and AI reading comprehension MRC42 to build an evidence-based classification and response system.

BNZ Partners led the development of classification criteria and the construction of training data. It incorporated a classification system aligned with K-Taxonomy compliance assessment procedures for green finance.

Kim Dong-hwan (김동환), chief executive of Fortytwo Maru, said the AI-based K-Taxonomy compliance assessment system has made standardisation and automation of external review processes for green finance a reality. He said it can lower entry barriers for mid-sized and small companies and provide a foundation to improve the speed and consistency of screening and disclosures. He said Fortytwo Maru will expand an ESG AI standard operating model that can be verified in practice through data quality management and continued model upgrades.