Constitutional AI: Realising Safe, Honest, and Trustworthy AI

Today, I returned to theUniversity of Tokyo ( UTokyo ) to study developments in Artificial Intelligence ( AI ) law. I have previously visited this renowned institution, including attending international AI conferences with participants from various countries. Established in 1877 as Japan’s first national university, UTokyo houses a Centre for Generative AI that is developing new approaches beyond current AI limitations. By combining human and artificial intelligence, the centre aims to advance scientific and technological progress through interdisciplinary collaboration across sciences, technology, and humanities, projected to benefit society, industry, economy, culture, and future research. It also shapes future societal visions and cultivates next-generation leaders. Japan is historically known for technological excellence while upholding human dignity and preserving traditions. UTokyo originally opened with four faculties: Law, Science, Literature, and Medicine. On Wednesday ( 27 May 2026 ), I visited the Law faculty library to access recent references on cyberlaw, AI, intellectual property, and personal data protection. One notable book I read was ’ Artificial Intelligence andTaxation Law: Double Taxation Dispute Resolution ’ by Enrico Gallo. Published for the first time in 2026, the book was available at the reception desk upon my arrival. I had pre-ordered it via email from Indonesia well before my Tokyo trip. The book identifies how AI systems can be integrated into decision-making processes for resolving international tax disputes.