Japan AI Safety Institute (AISI)
Material regulatory developments from Japan AI Safety Institute are tracked daily by Cresthaven Analytics. Over the last 90 days, 9 material briefs relevant to professionals operating in the technology, ai & competition sector have been published below. Each brief is sourced directly from the primary regulatory feed, summarized for institutional readers, and scored for materiality.
Recent material activity
A selection of recent published briefs; this is not a complete archive.
Japan and South Korea AI Safety Institutes hold inaugural joint technical workshop on AI safety evaluation
Japan's AI Safety Institute (AISI) and South Korea's AI Safety Institute held a two-day joint technical workshop on October 8 and 9, 2026. The session covered AI safety evaluation methodologies, multilingual and multicul…
Read the full Japan AI Safety Institute (AISI) brief →Japan AISI convenes 2026 general assembly of AI Safety Evaluation Environment Task Force with nine private-sector members
Japan's AI Safety Institute (AISI) convened the 2026 annual general assembly of its AI Safety Evaluation Environment Review Task Force on October 7, 2026, drawing approximately 120 attendees including remote public parti…
Read the full Japan AI Safety Institute (AISI) brief →Japan AISI publishes updated AI safety evaluation vision paper covering healthcare, robotics, education, and geospatial sectors
Japan's AI Safety Institute (AISI) released version 2.0 of its Business Demonstration Working Group Vision Paper on October 6, 2026. The update adds a new sub-working group and extends the shared evaluation framework to …
Read the full Japan AI Safety Institute (AISI) brief →Japan AISI publishes cyber capability evaluation of Claude Opus 4.8 using ExploitBench exploit benchmark
Japan's AI Safety Institute published an experiment note on October 2, 2026 reporting results from its ExploitBench-based cyber capability evaluation of Anthropic's Claude Opus 4.8. The evaluation was conducted in June 2…
Read the full Japan AI Safety Institute (AISI) brief →Japan AISI publishes experiment notes evaluating advanced AI cyber capability proliferation to open-weight models
Japan's AI Safety Institute published experiment notes, based on an evaluation conducted in July 2026, assessing whether advanced cyber capabilities found in frontier AI models have proliferated to publicly available ope…
Read the full Japan AI Safety Institute (AISI) brief →Japan AISI releases updated red teaming methodology guide for AI safety evaluation, version 1.10
Japan's AI Safety Institute published version 1.10 of its red teaming methodology guide in March 2025. The update adds concrete implementation examples, including a detailed walkthrough of red teaming applied to a Retrie…
Read the full Japan AI Safety Institute (AISI) brief →Japan AI Safety Institute updates AI safety evaluation guide to address proliferation of agent systems
Japan's AI Safety Institute published Version 1.20 of its Guide to Evaluation Perspectives on AI Safety in July 2026. The revision extends the guide's scope to cover AI agent systems, adding new evaluation perspectives a…
Read the full Japan AI Safety Institute (AISI) brief →Japan AI Safety Institute publishes second edition of known AI system attack taxonomy for red-team and security practitioners
On April 24, 2026, the Japan AI Safety Institute released the second edition of its reference document on known attacks and their impacts on AI systems. The update incorporates findings from recent international conferen…
Read the full Japan AI Safety Institute (AISI) brief →Japan AISI formalizes multilateral AI safety coordination framework spanning G7, OECD, and bilateral technical exchanges
Japan's AI Safety Institute has published a structured account of its international coordination activities. The release covers bilateral exchanges with counterpart AISIs in the United States and United Kingdom, multilat…
Read the full Japan AI Safety Institute (AISI) brief →
Get every Japan AI Safety Institute (AISI) brief delivered the day it lands
From $149/month
Cresthaven Analytics monitors Japan AI Safety Institute continuously, applying institutional materiality scoring to every release. Subscribers receive structured briefs via email, portal, and (Professional tier and above) real-time alerts.
View all tiers →