Anthropic releases new large model with excellent cybersecurity and vulnerability discovery capabilities

WealthCoffee · 2026-04-08T08:44:24+00:00

American startup Anthropic has released the Claude Mythos Preview, claiming that it can autonomously discover and exploit complex security vulnerabilities in major operating systems and browsers, with higher automation and success rates.

WealthCoffee

2026-04-08 08:44:24

Abstract generation in progress

**　　【Caixin Net】** Anthropic releases its “strongest network security AI model.” On April 7 local time in the United States, the U.S. AI model startup Anthropic released its latest large model, Claude Mythos Preview, claiming that it can independently discover and exploit complex vulnerabilities in major mainstream operating systems and web browsers without any human intervention.

Anthropic had previously released multiple AI models, and compared with its earlier versions, Claude Mythos Preview focuses on network security and vulnerability discovery capabilities. It can build complex exploit chains, and its level of automation and success rate have also shown significant improvements.

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.