Choose Rich Live
Tech

Japan asks OpenAI and Anthropic for expanded access to advanced AI models for cyber defense

Tokyo has requested greater access to frontier AI models from OpenAI and Anthropic to detect and patch system vulnerabilities, according to reports citing Nikkei.

Graphic titled 'Japan Seeks Frontier AI Access for Cyber Defense' detailing requests to OpenAI and Anthropic for Claude Mythos and GPT-5.5-Cyber models.
Image: @AGTPinsights

Japan has asked OpenAI and Anthropic for expanded access to their most advanced artificial intelligence models to bolster its cybersecurity, according to reporting from Nikkei. The Japanese government is seeking frontier models to identify and remediate vulnerabilities across its systems before malicious hackers can exploit them.[5][4][2][1][3]

Tokyo reportedly plans to seek access to Anthropic's Claude Mythos and OpenAI's GPT-5.5-Cyber. OpenAI had previously been weighing whether to provide GPT-5.5-Cyber, a model equipped with advanced cybersecurity capabilities. In addition to testing government systems for weaknesses, Japan also wants power and telecommunications providers to deploy the models.[2]

The outreach forms part of broader cybersecurity cooperation with the United States. Japan plans to feature deeper collaboration with American AI developers in an upcoming cybersecurity package scheduled for release as early as next week. No formal agreement between the Japanese government and either AI developer has been announced.[2]

Key facts

  • Nikkei reported that Japan asked OpenAI and Anthropic for expanded access to their most advanced AI models for cybersecurity purposes.
  • The specific models targeted in reports are Anthropic's Claude Mythos and OpenAI's GPT-5.5-Cyber.
  • Japan aims to use the AI models to run vulnerability checks on government systems and intends for power and telecom companies to use them as well.
  • The initiative is tied to closer U.S. cybersecurity cooperation and is expected to be included in a cyber package arriving as early as next week.
  • No official agreement between Japan and the two AI developers has been announced yet.

Sources · 4 sources

  1. CB

    Crypto BriefingArticle ·

    Japan seeks expanded access to advanced AI models from Anthropic and OpenAI Japan's pursuit of advanced AI access highlights its strategic focus on cybersecurity and international cooperation, impacting global AI dynamics. The post Japan seeks expanded access to advanced AI models from Anthropic and OpenAI appeared first on Crypto Briefing .

    Open source
  2. AG

    AGTP@AGTPinsightsPost on X ·

    Japan just asked OpenAI and Anthropic for more access to their most advanced AI models. Here's what you need to know. Nikkei reports Tokyo wants greater access to frontier models to find and fix cybersecurity vulnerabilities before hackers do. Hackers are getting more powerful AI tools, and Japan wants the same. The web research points to specific models. Japan reportedly plans to ask for expanded access to Anthropic's Claude Mythos and OpenAI's GPT-5.5-Cyber. OpenAI had earlier been weighing providing GPT-5.5-Cyber, which has advanced cybersecurity capabilities. The goal is to run vulnerability checks on government systems. Reports also say Japan wants power and telecom companies to use the models too. The request is part of closer cybersecurity cooperation with the US. The government plans to include stronger cooperation with US AI developers in a cybersecurity package due out as early as next week. Reactions split. Some say frontier models are now strategic assets, and countries without their own have to rely on others for defense. Others point to Japan facing near-daily AI-driven cyberattacks and say asking for the best models makes sense. Some critics frame it as Japan being dependent on American AI providers. Key numbers: - 2 companies asked: OpenAI and Anthropic - 2 models named: Claude Mythos and GPT-5.5-Cyber - Cyber package due as early as next week No agreement has been announced yet, and the request comes via Nikkei reporting.

    Open source
  3. CB

    Crypto BriefingArticle ·

    Japan seeks expanded access to Anthropic, OpenAI AI models for cybersecurity Japan's expanded AI access could strengthen its cybersecurity, impacting global AI collaboration and boosting Anthropic's market position. The post Japan seeks expanded access to Anthropic, OpenAI AI models for cybersecurity appeared first on Crypto Briefing .

    Open source
  4. *B

    *Walter Bloomberg@DeItaonePost on X ·

    JAPAN ASKS ANTHROPIC, OPENAI FOR MORE ACCESS TO ADVANCED MODELS - NIKKEI

    Open source
  5. NA

    Nikkei Asia@NikkeiAsiaPost on X ·

    Japan asks Anthropic, OpenAI for more access to advanced models https://t.co/pbDmw8Oy8g

    Open source