中/EN
— reading now
Flash

Flash·Safety & Risk·2026-10-10 02:06

Anthropic Launches Free AI-Powered OSS Scanner for Open-Source Projects

Anthropic on Thursday launched OSS Scanner, a free service that uses its most capable models (including Claude Mythos) to produce AI-generated vulnerability reports for eligible open-source projects — no human review required. The company says it has already flagged more than 29,000 candidate vulnerabilities in major open-source projects worldwide.

HIMIAO Horizon·2026-10-10 02:06·1 min read

Listen Save Share card Copy link
Magnifying glass over a screen showing code and vulnerability reports in a bright office, with shield security motifs
Opening its strongest models to open source for free: Anthropic's new move in defensive AI security.

Related

  1. 01

    Anthropic Launches Claude Haiku 5.5: From $0.10, About 75% Cheaper Than Its Predecessor

  2. 02

    Anthropic Cuts Live Internet Access for All Internal Evals

  3. 03

    Anthropic's New Rule: Sustained, Needless Abuse of Claude Can Lead to Account Restrictions — Effective Nov 12

  4. 04

    Anthropic Expands Cyber Verification Program: After 129,000 Bugs Found, Its Strongest Models Open to Vetted Defenders — With Fewer Safeguards

What happened

OSS Scanner is open to eligible open-source projects, whose core maintainers can enroll by submitting a GitHub PR. Scan reports are generated fully automatically by AI, covering common vulnerability classes with fix prioritization. As of October 9, 116 PRs had been submitted, including applications from crypto projects Nethermind and ZEUS on Friday.

Key facts

  1. The service:free, recurring AI-generated vulnerability scan reports for open-source projects — no human review required.
  2. Model power:reportedly using Anthropic's most capable models, including Claude Mythos.
  3. How to join:core maintainers enroll via GitHub PR; 116 sign-up PRs as of October 9.
  4. Results so far:per Anthropic, over 29,000 candidate vulnerabilities found in major open-source projects worldwide.

Why it matters

The service is free and uses models including Claude Mythos to generate vulnerability reports with no human review; Anthropic says it has flagged more than 29,000 candidate vulnerabilities, with 116 sign-up PRs as of October 9.
Useful Tap if this story helped you

CompaniesAnthropic

SourcesThe Hacker News (2026-10-09, source); iTechPost (2026-10-09, source). This article is a compilation of public information and does not constitute investment advice.

ANTHROPICCYBERSECURITYOPEN SOURCE

Comments

  1. Loading comments…

Today

Oct 12 Monday

Latest news

All →
  1. Oct 11GSK Expands Chai Deal After Wet-Lab Validation of AI Designs
  2. Oct 11PPT Master Hits GitHub Trending: Documents Become Native PowerPoint
  3. Oct 11context-mode Hits GitHub Trending: Tool Output, Sandboxed First
  4. Oct 11Cloudflare Acquires Deno: Deploy Shuts Down in Six Months
  5. Oct 11Anthropic Updates Claude Usage Policy: Armed Drones Named and Banned, Effective Nov 12
  6. Oct 11Google's $15 Billion India AI Data Center Sparks Backlash; Company Responds to Water and Power Fears
Ask the cat