Safety · 2026-10-08

Anthropic launches service that searches for vulnerabilities in open-source software

Anthropic, the company that develops the AI model Claude, is offering a service for finding security flaws in open-source programs, and participation is possible only on a voluntary basis.

✨ What we know about the news

Anthropic has announced the launch of a service that helps find vulnerabilities in open-source software. This term refers to programs whose code can be viewed and changed by anyone, and they are often created by volunteers or small teams.

Participation in the service is voluntary: developers decide for themselves whether to submit their projects for review.

📌 Why this matters to ordinary people

Many programs we use every day are partially built on open-source code. If such a program has a hidden security flaw, it can lead to problems with data or with how services work.

When such flaws are found and fixed before criminals can exploit them, users face less risk. This is exactly what the new service is aimed at.

More detailed terms, ways of working, and the list of partners are available in Anthropic's original announcement.

Sources: 🔗 Google News