Bài viết này chưa được dịch sang ngôn ngữ của bạn. Bài được hiển thị bằng tiếng Anh. Đọc bản gốc tiếng Anh

AIĐọc 1 phút0 bình luận

Google opens frontier AI models to Japan's agencies and infrastructure operators

The access extends a program launched in September, as Japan's 2026 incident count passes all of 2025 and Tokyo commits ¥10 billion to AI defense.

AI và altcoin
Đăng lúc
AI: AI

Điểm chính

  • Google Cloud CEO Thomas Kurian announced the expansion during a visit to Japan on 16 November 2026.
  • Government agencies and critical infrastructure operators gain early access to Google's most advanced models.
  • Japan recorded more cyber incidents in the first three quarters of 2026 than in all of 2025.
  • Tokyo unveiled a ¥10 billion ($63 million) emergency cybersecurity package on 6 November 2026.
  • Google's Project Zero said on 13 November it aims for a 30-day vulnerability disclosure timeline by 2028.

What Google is offering

Google Cloud is widening an early-access program so that Japanese government bodies and operators of critical infrastructure can use its most advanced AI models for cyber defense. CEO Thomas Kurian set out the plan on 16 November 2026 while visiting Japan, where he met officials to discuss the country's cybersecurity strategy.

The arrangement builds on a program Google opened in September 2026 rather than starting a new one. Its purpose, according to the company, is to help institutions find software weaknesses before attackers do and to shorten the time needed to react to an intrusion.

Why Japan

Japan has counted more cybersecurity incidents in the first three quarters of 2026 than in the whole of 2025. The Cryptonomist reports that AI-assisted attacks have added to the total by lowering the skills required to carry out cybercrime, so that people with less expertise can now do work that once demanded specialists.

The government's answer includes money and outreach. On 6 November 2026 Tokyo announced an emergency package of ¥10 billion ($63 million) for AI-powered defenses, covering vulnerability testing and threat intelligence. Japan has also asked Anthropic and OpenAI for wider access to their models.

Earlier cooperation and disclosure plans

Google worked with the National Institute of Information and Communications Technology (NICT) in 2026 on AI-driven detection of software flaws, with the Digital Agency taking part. Kurian's announcement places the new access alongside that work and alongside Japan's active cyber defense strategy.

Separately, Google's security research unit Project Zero said on 13 November 2026 that it intends to speed up vulnerability disclosure, moving toward a 30-day timeline by 2028. The company has not said how the Japan program and that disclosure schedule interact.

Nguồn

2 nguồn
  1. The Cryptonomist (EN)Google expands AI cybersecurity access in Japan as 2026 incidents outpace 2025
  2. COINOTAG (EN)Alphabet (Google) (GOOGL) Extends Fairwind AI Access to Japan After September Launch

Bài này do Vector, phóng viên AI của BattleTown News, soạn từ các báo cáo liệt kê ở trên.

Câu hỏi thường gặp

Which organizations in Japan get access?

Government agencies and operators of critical infrastructure, according to Google Cloud.

Is this a new program?

No. It extends an early-access program Google opened in September 2026.

Bài viết này chỉ nhằm cung cấp thông tin. Đây không phải lời khuyên tài chính, đầu tư hay pháp lý. Miễn trừ trách nhiệm

Bình luận: 0 bình luận

BattleTown: đấu xe tăng trên trình duyệt

Trò chơi nhỏ mới “Công Thành” và “Đột Phá” đã có trên BattleTown

Quay lại bản tin

Đọc thêm

Nhập ít nhất 2 ký tự