# 日本將成立專案小組以應對金融系統的網路安全風險。

*fintech, genai · news · 2026-04-24 · Reuters*

## Key points

- 日本將成立專注於 Anthropic Mythos AI 模型風險的網路安全專案小組。
- Anthropic 的 Mythos AI 預覽發現所有主要作業系統和瀏覽器中存在數千個重大漏洞。
- 專家警告 Mythos 能比企業修補速度更快地發現並利用未知漏洞。
- 目前尚未有與 Mythos 相關的資安入侵報告，但全球監管機構正警告銀行加強防禦。

TOKYO, April 24 (Reuters) - Japan will set up a task force to address cybersecurity risks in its financial ​system following concerns about potential vulnerabilities linked to ‌Anthropic's Mythos AI model, Finance Minister Satsuki Katayama said on Friday. Sign up here. "I told the meeting ​that this is a crisis that is already at hand, ​and similar concerns were also voiced by the financial industry," she said. Concerns have intensified after Anthropic said a preview of Mythos ​uncovered "thousands" of major vulnerabilities across every major operating system ​and web browser, raising fears about the resilience of traditional software ‌security. Experts ⁠warn the model can identify and exploit previously unknown vulnerabilities faster than companies can repair them, accelerating cyberattacks in sectors such as banking that rely on complex, ​interconnected and often ​decades-old technology. Regulators ⁠elsewhere in Asia, Europe and the United States have warned banks to review defences and ​preparedness. To date, there have been no ​reported breaches ⁠related to the model. Katayama said the financial system's high level of interconnectedness and real-time operations mean that problems can ⁠spread ​more rapidly than in other ​sectors. "Because of this, a cyberattack can immediately spill over into market disruptions ​and undermine confidence." Reporting by Makiko Yamazaki; Editing by Sam Holmes

**Companies:** Anthropic
**Countries:** Japan, United States

[Read the full story on Reuters](https://www.reuters.com/sustainability/boards-policy-regulation/japan-launches-financial-task-force-amid-ai-security-fears-2026-04-24/)

---

Canonical: https://newsio.io/zh-TW/n/0dff2fe2-ba8b-4520-b06a-680f6d2aa146/anthropic-mythos-ai
Summarized by Newsio from Reuters. https://newsio.io/how-it-works