[2M4-GS-11a] AI and Society

1:45 PM - 2:00 PM JST(4:45 AM - 5:00 AM UTC)

[2M4-GS-11a-02] Public Disclosure of AI Ethics Policies in Japanese Corporations: A Survey of Major Companies

〇Kumiko Komatsu¹, Nina Ždanovič¹, Ryuma Shineha¹, Shutaro Takeda¹ (1. Keio University)

Keywords:

AI ethics,corporate governance,AI governance,AI policy,public disclosure

This study clarifies the public disclosure status and issue emphasis of stand-alone AI governance or AI ethics policies among major Japanese corporations. Using a revenue-based sample of large firms drawn from a commercial database, we conducted a systematic web-based document survey to identify whether each company publishes an externally accessible policy on its official website. We find that public disclosure remains limited: only 47 of 337 firms (13.9%) disclose at least one stand-alone AI ethics document, while more than 80% show no evidence of external disclosure, with substantial variation across sectors. We then examine issue coverage within the disclosed documents (n = 47) using a dictionary-based coding scheme and observe a strong concentration of emphasized topics: privacy, transparency/explainability, governance and regulation, safety, and fairness/bias are addressed by almost all disclosing firms, whereas writing/research, cybercrime, interaction risk, hallucinations, and art/creativity appear in only a small subset of corporate disclosures.