Skip to content
Visa Open-Sources AI Tool After Identifying 10,000 Vulnerabilities

Visa Open-Sources AI Tool After Identifying 10,000 Vulnerabilities

First seen 30 Sep 2026, 21:37 UTC • •
API

Article Content

Browse articles
ThreatCluster AI
ThreatCluster •September 30, 2026 at 22:31 UTC
  • •Visa has open-sourced its AI tool to combat over 10,000 identified vulnerabilities.
  • •The Visa Vulnerability Agentic Harness employs AI to discover and fix security flaws.
  • •Rajat Taneja warns of the growing threat from autonomous cyberattacks.

Visa has open-sourced its AI-based framework, the Visa Vulnerability Agentic Harness (VVAH), to address over 10,000 vulnerabilities identified by Anthropic's Mythos AI model. The VVAH is designed to discover, verify, and fix security flaws in software while maintaining human oversight. This initiative comes amid growing concerns over the risks posed by advanced AI and potential autonomous cyberattacks. Visa processes around one billion payments daily, making its security critical to the global financial system. Rajat Taneja, Visa's President of Technology, emphasized the need for agentic defenses in response to evolving threats. The tool is available on GitHub under an Apache 2.0 license, requiring Python 3.11 and an Anthropic API key for operation. The vulnerabilities identified were of high severity, highlighting the urgency for improved security measures.

Start a free Starter trial for enhanced analysis

Ask AI about this cluster

Updated just now How this analysis works

Timeline

2026-09-30
Visa open-sources VVAH
Visa released the Visa Vulnerability Agentic Harness to address vulnerabilities found by Anthropic's Mythos AI model.
Pasqualepillitteri.It
Recent
10,000 vulnerabilities identified
Anthropic's Mythos AI model identified over 10,000 high or critical severity vulnerabilities in Visa's software.
Computing

More articles in this cluster (2)

Extracted Entities (12)

1 / 2

Common questions

What vulnerabilities were identified?
Over 10,000 high or critical severity vulnerabilities were identified by Anthropic's Mythos AI model.
How does the VVAH work?
The VVAH uses AI agents to discover, verify, and fix vulnerabilities while ensuring human oversight.
What is the significance of this release?
The open-sourcing of VVAH aims to enhance security measures against evolving cyber threats.