Regulation
Neutral
Antropic Expands Security Program After Discovering 129,000 Vulnerabilities

Antropic is expanding its security program to allow certified cybersecurity professionals to use its high-performance artificial intelligence (AI) models with certain safety measures relaxed. The move follows a project this year that identified over 129,000 verified vulnerabilities in core software. According to Reuters, Antropic announced on June 6 (local time) the integration and revamp of its existing 'Project Glasswing' and Cyber Verification Program (CVP). The new program will categorize initiatives based on verification levels and security controls.
Korean Source
This article is an English localization of a Korean-language crypto news report. Original headline: AI로 취약점 12만9000건 찾았다…앤트로픽 보안 프로그램 확대