Japan AI Safety Institute (J-AISI)

Japan's AI safety institute.

Category

Governance

Active since

Feb 2024

Safe Singularity

Japanese-language X account on superintelligence risk and alignment, run by AGI Hub's Chief Alignment Officer.

Category

Advocacy, Blog

Active since

May 2026

AGI Hub

Tokyo organisation communicating AGI's societal impact to a Japanese audience — lectures, explainers, translations, and the AI 2040 scenario project. Founded June 2026.

Category

Advocacy, Training and education

Active since

Jun 2026

bioshok

AI safety writer covering alignment, governance, x-risk, and superintelligence.

Category

Blog

Active since

Jan 2023

AIST AI Quality & AI Safety

AI quality and safety programs at AIST (National Institute of Advanced Industrial Science and Technology, 産総研). Develops evaluation and management technology for AI quality and safety in tandem with rule-making — including the AI Quality Management (AIQM) project, the Machine Learning Quality Management Guideline, and the NEDO-commissioned AI Safety Enhancement R&D program.

Category

Empirical research

Shiba AI Lab

Tokyo-based research lab working on AI safety. Themes include evolutionary approaches to AI development, multi-agent coordination, and interpretable AI systems.

Category

Empirical research

Active since

Feb 2026

AI Safety Tokyo

Tokyo-based AI safety community running regular benkyoukai (study group) events and reading groups.

Category

Training and education

Active since

Jan 2023

TARA

APAC-wide technical AI safety training program. 14-week ARENA curriculum via Saturday sessions; Tokyo is one of six 2026 cohort cities.

Category

Training and education

Active since

Mar 2026

AI Alignment Network (ALIGN)

Japan's AI alignment research nonprofit. Original research (Post-Singularity Symbiosis, machine ethics), webinars, study groups, and outreach across academia and industry.

Category

Training and education, Conceptual research

Active since

Sep 2023

ExtinctionMeme

Japanese-language AI safety meme account — translates and posts x-risk memes to lower p(doom). Handle name is the JP title of If Anyone Builds It, Everyone Dies.

Category

Advocacy

Active since

Aug 2026

Artificial Life Institute

Kyoto-based nonprofit researching evolutionary AI at the intersection of computer science and evolution. Explores alternative pathways to intelligence and symbiosis between different forms of intelligence.

Category

Conceptual research

Noeon Research

Tokyo-based deep-tech research lab developing a novel AI architecture for general reasoning grounded in category theory and graph-based knowledge representation. Pursues AGI via interpretable, alignment-friendly foundations rather than black-box neural networks.

Category

Conceptual research

Active since

Jan 2020

Whole Brain Architecture Initiative (WBAI)

Tokyo-based nonprofit advancing brain-inspired artificial general intelligence through the Whole Brain Architecture approach. Develops software platforms for AGI research and connects researchers through seminars, workshops, and hackathons.

Category

Conceptual research

Active since

Aug 2015

UTokyo AI Safety (UAS)

AI safety student circle at the University of Tokyo. 東京大学のAIセーフティ学生サークル。

Category

Training and education