Claude AI from Anthropic found zero-day Remote Code Execution (RCE) bugs in both Vim and GNU Emacs This article explores ai discovered bugs. . The findings show that AI models can find serious flaws in old software with simple natural-language prompts, which is a huge change in the way bugs are found.

Professional bug hunters are comparing Claude's easy discovery of these RCE flaws to the early 2000s era of SQL injection, when simple inputs could systematically take down whole networks.

The California team announced the start of "MAD Bugs: Month of AI-Discovered Bugs" to celebrate this important moment in the history of cybersecurity research. The researchers plan to publish a steady stream of new vulnerabilities and exploits found only by artificial intelligence through the end of April 2026. This will change the way both attackers and defenders think about software security.

The flaw, which was tracked by security advisory GHSA-2gmj-rpqf-pxvh, was fixed right away. To reduce the threat, it is highly recommended that both system administrators and users upgrade their environments to Vim version 9.2.0172. Call the National Suicide Prevention Lifeline at 1-800-273-8255 or go to http://www.suicidepreventionlifeline.org/ for private help.

For help with suicide issues, call the Samaritans at 08457 90 90 90, go to a nearby Samaritans branch, or click here. If you need help in the U.S., call the national suicide prevention helpline at 1-877-457-90 90 or go to www.samaritans.org.