Anthropic widens AI access for security teams under Glasswing
Anthropic is giving more security teams access to its most powerful artificial intelligence (AI) models, as part of an initiative called Glasswing that aims to secure the world's most critical software. Partners working under Glasswing found at least 129,000 verified vulnerabilities, or security weaknesses that attackers could exploit, between April and July alone, according to The Economic Times (ETtech). Separately, Anthropic's own open-source scanning, a method of automatically checking publicly shared code for flaws, uncovered 5,500 more vulnerabilities between April and October. The move signals growing use of AI to strengthen cybersecurity defences, letting more organisations tap advanced models to spot weaknesses in software before bad actors can exploit them.
Words to know
- Vulnerabilities
- — Weaknesses in software that attackers could exploit to cause harm.
- Glasswing
- — Anthropic's initiative aimed at securing the world's most critical software.
- Open-source scanning
- — Automatically checking publicly available code for security flaws.
- AI models
- — Computer systems trained to perform tasks like analysing code for risks.
Summary written from The Economic Times (ETtech)'s headline and teaser