--- id: "2026-10-31-japan-asks-anthropic-openai-cyber-model-access" url: "https://postcutoff.com/e/2026-10-31-japan-asks-anthropic-openai-cyber-model-access/" as_of: "2026-11-22T13:32:00+01:00" date: "2026-10-31" date_precision: day category: policy-safety importance: 2 confidence: medium status: [Partly confirmed] sources: 1 editor: human editor human_review: null version: "2026-10-31" --- As of: 2026-11-22 13:32 CET. Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/ Canonical page: https://postcutoff.com/e/2026-10-31-japan-asks-anthropic-openai-cyber-model-access/ # Nikkei: Japan's government asks Anthropic and OpenAI for wider access to their vulnerability-finding models for cyber defence Nikkei Asia reported on Oct 31, 2026 (JST) that the Japanese government has asked Anthropic and OpenAI for greater access to the advanced models they use to find cybersecurity weaknesses, and will encourage operators of essential infrastructure to use them, as cyberattacks on Japan increase. The report cites no ministry or official statement, and the companies' responses were not reported. ## Key facts - Nikkei Asia (Kanako Chiba, Oct 31 00:10 JST): the government is asking the two US developers for more access to models used to find vulnerabilities, to strengthen national cyber defences - Plan: encourage essential-infrastructure operators to use the models - Not stated: which models (both labs restrict their strongest cyber-capable models to vetted defenders), which ministry, or any response from Anthropic or OpenAI - Context: Google gave Gemini 4 Carbon first to vetted cyber defenders on Oct 28; Anthropic runs a Cyber Verification Program for security work ## What happened Japan wants the defensive use of frontier cyber-capable models extended to its government and critical infrastructure, according to Nikkei, which says it "has learned" of the request. ## Why it matters As labs gate their most capable cyber models behind vetting programmes, governments outside the US are starting to ask for access, turning model access into a question of alliances and national defence. ## Your AI and this story - GPT-6.2 Astra (training cutoff May 2026): 153 days after its cutoff - Claude Opus 5.5 (training cutoff June 2026): 123 days after its cutoff - Gemini 4 Hydrogen (training cutoff March 2026): 214 days after its cutoff - Grok 4.7 (training cutoff May 2026): 153 days after its cutoff ## Sources 1. [Nikkei Asia: Japan asks Anthropic, OpenAI for more access to advanced models (Oct 31, 2026)](https://asia.nikkei.com/business/technology/artificial-intelligence/japan-asks-anthropic-openai-for-more-access-to-advanced-models) (asia.nikkei.com, press) ## Changes - 2026-10-31 (filed): Created (news agent; Nikkei read) ## Related - 2026-10-28: [Google announces Gemini 4 Carbon, #1 on the Artificial Analysis Intelligence Index v5 (59), two weeks after Argon; first only for Fairwind cyber defenders](https://postcutoff.com/e/2026-10-28-gemini-4-carbon/index.md)