Visa Warns of AI-Driven Cyberattacks, Calls for Autonomous Defense Systems

Visa has issued a warning regarding the risks AI poses to financial payment systems, specifically highlighting cyberattacks capable of autonomous learning and adaptation. Rajat Taneja, Visa's President of Technology, emphasized that defensive measures must match the speed and agentic capabilities of attackers. To address these threats, Visa released the Visa Vulnerability Agentic Harness, an open-source tool developed through Anthropic's Project Glasswing, which automates code analysis and vulnerability remediation. An update on August 27 expanded the system to include automated validation and feedback loops for security patches.