AI Model Analyzes Corporate Sustainability Priorities

Kyushu University researchers developed a multilingual AI model, Text Match Pre-Trained Transformer (TMPT), to assess how corporate disclosures align with financially material ESG issues. By analyzing 1,623 annual reports from 340 Japanese food and beverage companies between 2020 and 2024, the study found firms prioritize social topics over environmental reporting.

Unlike simple keyword counting, TMPT uses zero-shot semantic matching to evaluate the depth of sustainability claims. The study, published September 16, 2026, in Corporate Social Responsibility and Environmental Management, demonstrates that corporate reporting priorities shift alongside changing regulatory requirements and industry-specific benchmarks established by the Sustainability Accounting Standards Board.

Editor's note: Accurate and concise summary of the research methodology and findings provided in the text.