Anthropic Releases GLM-5.3 with End-to-End Network Utilization Capabilities

On September 29, 2026, Anthropic Frontier Red Team published a research paper stating that the GLM-5.3 developed by Anthropic AI possesses autonomous end-to-end network utilization capabilities. About five months ago, Anthropic released the Claude Mythos Preview through Project Glasswing, discovering over 10,000 vulnerabilities. GLM-5.3 is released with open weights and lacks abuse restrictions; in simulated tests, the success rate of bypassing its protections using simple methods ranged from 64% to 100%. The National Institute of Standards and Technology assessed that GLM-5.3 is the most powerful open-weight model for network capabilities to date, with capability conclusions consistent with the frontier, lagging approximately four months behind.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.